UC Irvine UC Irvine Previously Published Works

Title Who is susceptible in three false memory tasks?

Permalink https://escholarship.org/uc/item/34d2k097

Journal Memory (Hove, England), 27(7)

ISSN 0965-8211

Authors Nichols, Rebecca M Loftus, Elizabeth F

Publication Date 2019-08-01

DOI 10.1080/09658211.2019.1611862

Peer reviewed

eScholarship.org Powered by the California Digital Library University of California Memory

ISSN: 0965-8211 (Print) 1464-0686 (Online) Journal homepage: https://www.tandfonline.com/loi/pmem20

Who is susceptible in three false memory tasks?

Rebecca M. Nichols & Elizabeth F. Loftus

To cite this article: Rebecca M. Nichols & Elizabeth F. Loftus (2019) Who is susceptible in three false memory tasks?, Memory, 27:7, 962-984, DOI: 10.1080/09658211.2019.1611862 To link to this article: https://doi.org/10.1080/09658211.2019.1611862

Published online: 02 May 2019.

Submit your article to this journal

Article views: 182

View Crossmark data

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=pmem20 MEMORY 2019, VOL. 27, NO. 7, 962–984 https://doi.org/10.1080/09658211.2019.1611862

Who is susceptible in three false memory tasks?

Rebecca M. Nicholsa and Elizabeth F. Loftusb aSchool of Social Sciences, Nanyang Technological University, Singapore, Singapore; bPsychology and Social Behavior, University of California, Irvine, CA, USA

ABSTRACT ARTICLE HISTORY Decades of research show that people are susceptible to developing false memories. But if Received 22 January 2019 they do so in one task, are they likely to do so in a different one? The answer: “No”.In Accepted 16 April 2019 the current research, a large number of participants took part in three well-established KEYWORDS false memory paradigms (a misinformation task, the Deese-Roediger-McDermott [DRM] list False memory; memory learning paradigm, and an imagination inflation exercise) as well as completed several ff distortion; misinformation; individual di erence measures. Results indicate that many correlations between false DRM; imagination inflation; memory variables in all three inter-paradigm comparisons are null, though some small, individual differences; false positive, significant correlations emerged. Moreover, very few individual difference variables memory susceptibility significantly correlated with false memories, and any significant correlations were rather small. It seems likely, therefore, that there is no false memory “trait”. In other words, no one type of person seems especially prone, or especially resilient, to the ubiquity of memory distortion.

As much as we would like to believe so, human memory is The misinformation effect not the foolproof gatekeeper of our past. Though our memories generally serve us well in our day-to-day lives Background and theory and contribute meaning to ourselves and the physical One commonly used paradigm involves exposure to misin- and social environments in which we are so deeply formation about a past memory (Loftus, Miller, & Burns, embedded, we do not encode and retrieve in a failsafe 1978). In a typical misinformation study, participants record-and-playback manner. Rather, our memory pro- witness some type of event and later are given some mis- cesses can be vulnerable to contamination and distortion. information (i.e., inconsistent post-event information) Our memories are malleable, susceptible to influence from about that event. When tested on their memory for suggestion, and can even contain whole events that never event details, participants often incorporate pieces of mis- actually occurred. information into their memory of the original event and The past four decades have seen an explosion of report those details as such. For example, participants in research investigating the psychological underpinnings one study viewed a slideshow in which a burglar picked of faulty memory (for a review of some of this work up a hammer, then sometime later read misleading infor- and its motivations, see Brainerd & Reyna, 2005 or mation that the burglar handled a screwdriver instead. At Loftus, 2017). This collective body of work has shown test, they were asked whether they saw a hammer or a that memories can be manipulated experimentally. This screwdriver in the original slideshow. Across experiments, holds for memories of simple events created in the lab participants selected the misinformation item – the screw- and memories for perceptually rich and personally mean- driver – 66% of the time, representing a typical pattern of ingful events from one’s past. What’s less known, memory impairment in misinformation studies (McCloskey however, is whether individuals who are particularly & Zaragoza, 1985). prone to developing false memories in one context are One of the predominant theoretical explanations for the also prone to developing phenomenologically different misinformation effect is that of errors in source monitoring false memories in other contexts. We examined this (Johnson, Hashtroudi, & Lindsay, 1993). The source moni- issue using three commonly used paradigms for demon- toring account states that when participants encode the strating memory malleability. We describe their methods, misinformation, they may not adequately encode the their theoretical bases, and individual differences that source of the post-event information and thus erroneously predict susceptibility below. attribute it to the original event (for reviews, see Belli &

CONTACT Rebecca M. Nichols [email protected] School of Social Sciences HSS-04-26, Nanyang Technological University, 48 Nanyang Avenue, Singapore 639818, Singapore This article has been republished with minor changes. These changes do not impact the academic content of the article. © 2019 Informa UK Limited, trading as Taylor & Francis Group MEMORY 963

Loftus, 1994; Mitchell & Johnson, 2000; Zaragoza, Lane, effect. For example, working memory has been shown to Ackil, & Chambers, 1997). This can easily happen because be negatively correlated with misinformation-related false the original information and misinformation share some memory (Calvillo, 2014; Jaschinski & Wentura, 2002). Rela- commonalities (e.g., the referent event and the context in tively low intelligence coupled with poor perceptual which those events occur) that inherently make source dis- capacity is also associated with susceptibility to misinfor- tinction difficult. Johnson et al. (1993) identified a number mation (Zhu et al., 2010a). More specifically, individuals of conditions in which source misattributions are particu- with a low degree of cognitive ability combined with larly likely, which include imagining perceptual detail at traits such as low fear of negative evaluation, low need the time of encoding misinformation, encoding misinfor- to avoid harm, high cooperativeness, high dependence mation that is particularly congruent with the overall on reward, and high self-directedness seemed to be meaning of the event, and experiencing stress or fatigue especially susceptible to the influence of misinformation during encoding of the misinformation. (Zhu et al., 2010b). It may be the case that a specific Another approach to explaining misinformation false threshold of cognitive ability is required both to accurately memories is fuzzy trace theory (FTT; Brainerd & Reyna, recall witnessed events and to discern the sources of 2005). FTT posits two types of mental representations for potentially competing information about those events. memory: verbatim traces, which are detailed, often vivid, direct representations of the past, and gist traces, which The Deese-Roediger-McDermott paradigm capture overall meaning of the memory but not its specific qualities. According to FTT, endorsement of misin- Background and theory formation is supported by both verbatim and gist trace Nearly twenty years after the misinformation effect was strength. An individual may misremember the item as a demonstrated, Roediger and McDermott rediscovered the screwdriver because they recall this verbatim information work of Deese (1959) and developed what became from the misinformation phase, or they may misremember another widely used experimental manipulation designed the item as a screwdriver because it is consistent with the to induce false memories, the Deese-Roediger-McDermott overall gist memory that the item was a tool. (or DRM) paradigm (Roediger & McDermott, 1995; Stadler, Roediger, & McDermott, 1999). In the DRM, participants Individual differences view or hear lists of semantically-related words such as Compared to the large corpus of research on the misinfor- note, sound, piano, sing, radio, band, etc., which converge mation effect, relatively few studies explore individual upon a critical, non-presented semantic associate (in this differences in susceptibility. Age is one of them; older case, music). Participants are tested on their memory for adults have shown to be more susceptible to memory studied items, new but unrelated control (distractor) items, errors, and this is likely because of decreased frontal lobe and of most interest, critical items. The rate of false mem- functioning and thus poorer source memory discriminabil- ories for critical items during encoding often approximates ity (Roediger & Geraci, 2007). Other research has shown or exceeds that of veridical memory for studied items, and that people who report stronger ability to imagine visual these patterns have been robustly replicated in the DRM lit- images are more likely to endorse misinformation (Cann erature (for reviews of the DRM, see Gallo, 2006; 2010). & Katz, 2005; Eisen, Gomes, Lorber, Perez, & Uchishiba, One theory that accounts for false memories in the DRM 2013; Tomes & Katz, 1997), though imagery ability has is the activation-monitoring hypothesis (Roediger & also been linked to a close confidence-accuracy relation- McDermott, 2000; Roediger, Watson, McDermott, & Gallo, ship in the misinformation paradigm (Tomes & Katz, 2001). This theory posits that the critical item is mentally 2000). Furthermore, a positive association has been generated both at encoding and potentially during retrie- found between subscales of the Openness and Agreeable- val as a result of spreading activation, a process by which ness dimensions of the NEO Personality Inventory and mis- nodes that represent concepts in a semantic network are information false memory (Liebman et al., 2002), which is activated due to the proximity and magnitude of activation consistent with the idea that social factors play a role in of other nodes (Underwood, 1965). Faulty monitoring the acceptance of misinformation (McCloskey & Zaragoza, during study then produces source confusions at test, 1985). Furthermore, the degree to which participants self- and participants mistakenly remember the critical item as report dissociative experiences has also been shown as a having been presented, instead of merely cognitively acti- predictor of incorporation of misinformation into vated, at study. FTT has also been proposed to explain DRM memory (Eisen & Carlson, 1998; Eisen, Morgan, & Mickes, memories; from an FTT perspective, false recognition of 2002; Hekkanen & McEvoy, 2002), which is not surprising music at test may occur because it is consistent with given that those who dissociate (particularly those with overall gist memory that the list contained items conver- pathological dissociation) are likely to suffer from a ging upon music (Cabeza & Lennartson, 2005). number of cognitive problems that include limits in source monitoring (Putnam, 1995). Individual differences In addition, cognitive factors seem to play a role in Relative to the misinformation effect, individual predictors understanding one’s susceptibility to the misinformation of susceptibility to false memories of critical lures in the 964 R. M. NICHOLS AND E. F. LOFTUS

DRM have been extensively studied. Just as with the misin- Another cognitive variable implicated in individual formation effect, age is negatively correlated with DRM difference research on the DRM is cognitive style, or false memories in adults (Balota et al., 1999; Butler, McDa- one’s preferred general approach to information proces- niel, Dornburg, Price, & Roediger, 2004; Gallo, Bell, Beier, & sing (see Sternberg, 1997, for a review). One way to Schacter, 2006; Kensinger & Schacter, 1999). With respect measure cognitive style is with the Need for Cognition to the NEO Personality Inventory, extraversion has been (NFC) scale, which measures the degree to which one shown to correlate with false memory susceptibility. prefers to engage in effortful thinking (Cacioppo & Petty, Specifically, extraverts have been shown to falsely remem- 1982). Individuals higher in NFC have been shown to be ber more critical lures than both introverts and ambiverts more susceptible to DRM errors in both tests of recognition (Sanford & Fisk, 2009). This finding is consistent with evi- (Graham, 2007) and recall (Leding, 2011; Wootan & Leding, dence that suggests extraversion is correlated with heigh- 2015). It is perhaps the case that those with greater NFC tened arousal, which in turn increases spreading activation engage in more elaborative cognitive processing or have and thus leads to greater effects of semantic priming (see more integrated semantic networks that allow for easier Matthews & Harley, 1993). false retrieval of nonpresented critical lures at test. There Furthermore, many studies have found a positive corre- is also evidence to suggest that one’s Faith in Intuition lation between dissociative experience and false memories (FI), a construct that is orthogonal to NFC in the Rational- in the DRM (Clancy, McNally, Schacter, Lenzenweger, & Experiential Inventory (see Epstein, Pacini, DenesRaj, & Pitman, 2002; Clancy, Schacter, McNally, & Pitman, 2000; Heier, 1996) and which gauges one’s reliance on intuitive Dehon, Bastin, & Larøi, 2008; Wilkinson & Hyman, 1998; thought, is positively correlated with false memories in Winograd, Peluso, & Glover, 1998; Zoellner, Foa, Brigidi, & the DRM (Nichols & Loftus, 2009). When music is presented Przeworski, 2000), though some studies have failed to at test, participants may experience an immediate and observe such a relationship (e.g., Bremner, Shobe, & Kihl- intuitive feeling that it had been presented during encod- strom, 2000; Geraerts, Smeets, Jelicic, van Heerden, & ing, and this may be an especially operant mechanism for Merckelbach, 2005; Platt, Lacey, Iobst, & Finkelman, 1998; false memory development in those who tend to rely on Wright, Startup, & Mathews, 2005). Positive correlations their “gut feelings” more often than others more generally. also exist between DRM false memories and measures of Lastly, it has also been shown that people with a field- delusional ideation (Dehon et al., 2008; Laws & Bhatt, dependent processing style, who rely on background and 2005) and fantasy proneness (Geraerts et al., 2005)as holistic information in order to process globally, are more well as imagery ability (Winograd et al., 1998). The discov- susceptible to false memories than their field-independent ery of these relationships suggest that an overall predispo- counterparts, who can more easily identify individual, con- sition toward moving thoughts from consciousness, or stituent elements apart from the whole (Corson, Verrier, & otherwise holding imaginative beliefs more generally, Bucic, 2009). This is an intuitive finding consistent with may be related to overall worse memory (and more specifi- implications from spreading activation theory; DRM items cally in this case, a greater likelihood of using similarity to themselves contribute to part of a whole (the relevant judge critical lures as having been actually encoded). list, which semantically converges upon the lure), and Researchers have also studied a number of cognitive those who are more likely to process those items holisti- variables as predictors of false memories in the DRM. A cally are more likely to endorse the lure. growing body of research has demonstrated a negative Finally, it is important to note briefly that the proneness correlation with working memory (Parker, Garry, Engle, to DRM errors itself is considered itself a stable individual Harper, & Clifasefi, 2008; Peters, Jelicic, Haas, & Merckel- difference; one study demonstrated a correlation of r bach, 2006; Peters, Jelicic, Verbeek, & Merckelbach, 2007). = .76 between false alarms to lures at Time 1 and those This relationship appears to be stronger under conditions of Time 2, two weeks later (Blair, Lenton, & Hastie, 2002). that call for a high degree of source monitoring, such as when participants are forewarned about the types of mem- Self-relevant false memories: imagination inflation ories errors the DRM tends to produce (Watson, Bunting, Poole, & Conway, 2005). On the other hand, individuals Background and theory who report a higher number of everyday mental errors Several studies have shown that people can develop false (or who report less cognitive efficiency) have demon- memories for whole autobiographical events. During the strated reduced susceptibility to endorsement of critical highly profiled repressed memory cases of the 1990s, lures (Raymaekers, Peters, Smeets, Abidi, & Merckelbach, when claimants purported to have had experienced but 2011), suggesting that those with a more optimistic view subsequently repressed unspeakable crimes such as ritual of their own cognitive efficiency may have a global ten- satanic abuse (see Bottoms & Davis, 1997), it became dency to make liberal memory judgments, which in turn clear that more work was needed to investigate whether leads them to endorse more critical lures. Though the or not memories for entire events could be implanted. DRM does produce memory errors, results such as these And so was born a new line of false memory research: illuminate the rationale to consider response biases in that of personally relevant false memories. Unlike their mis- addition to (not instead of) memory errors. information and DRM counterparts, these memories may MEMORY 965 be comprised of whole, richly detailed events of one’sown incorporate detailed sensory elaborations during imagin- personal past (see Loftus & Bernstein, 2005 for a review). ation and not when imagined events lack sensory cues or One of the first studies to implant false autobiographical are otherwise not imagined with such perceptual precision memories did so with the lost-in-the-mall technique (Thomas, Bulevich, & Loftus, 2003). Thus, it is possible that (Loftus & Pickrell, 1995). In this study, participants were feelings of familiarity can facilitate the imagination inflation given short narratives of events that occurred in the past, effect, but it seems likely that source misattributions and for all participants in the study, one of the stories account for inflation over and above explanations related was about having gotten lost and subsequently found by to increased general familiarity or ease of processing. a family member in a shopping mall during childhood. One criticism of imagination inflation studies is that the Unbeknownst to the participants, family members had pre- imagination exercise actually allows the participant to viously verified this event never to have happened for any access a previously inaccessible yet true event from one’s of the participants. In the original study, a sizeable minority childhood, and the increase in confidence is an artifact of (i.e., 25%) of participants remembered the fictitious lost-in- this process (Pezdek & Salim, 2011). One experimental the-mall event after being subjected to a suggestive variant of the imagination inflation paradigm addresses interview. this concern. In this modified paradigm, adapted from Since then, self-relevant false memories have been Goff and Roediger (1998), participants come into the lab implanted using a variety of experimental techniques, and hear a list of actions, some of which are everyday including dream interpretation (Loftus & Mazzoni, 1998; tasks (e.g., flip a coin) and some of which are bizarre (e.g., Mazzoni, Lombardo, Malvagia, & Loftus, 1999), guided visu- kiss the magnifying glass). At a later time, they return and alisation (Paddock, Terranova, Kwok, & Halpern, 2000), hyp- imagine some of the performed events as well as some nosis (Spanos, 1996; Spanos, Burgess, Burgess, Samuels, & novel events. Finally, participants return later for a third Blois, 1999), false feedback (Bernstein, Laney, Morris, & and final time in which their memory is tested for the Loftus, 2005a, 2005b; Laney, Morris, Bernstein, Wakefield, actions performed in Session 1. One study that utilised & Loftus, 2008; Laney & Loftus, 2008), doctored photo- these methods found that repeated imagination led graphs (Strange, Sutherland, & Garry, 2006; Wade, Garry, people to report that both everyday and bizarre actions Don Read, & Lindsay, 2002), and doctored video (Nash & were performed during Session 1 when in fact they were Wade, 2009; Nash, Wade, & Lindsay, 2009). Another tech- merely presented, imagined, or even neither of the two nique used to implant false memories is guided imagin- (Thomas & Loftus, 2002). Because the referent events ation (Garry, Manning, Loftus, & Sherman, 1996). In a were controlled by the experimenters, the research directly typical guided imagination study, participants are given a addressed the issue concerning the potential access of true number of childhood events and are asked to rate their memories and confirmed imagination indeed inflates false confidence that they experienced each event as a child. memory. Participants are later asked to imagine one of the events and, sometime later, fill out another identical life events Individual differences inventory. An increase in confidence from the first inven- Few studies have examined individual differences in sus- tory to the second for the critical event (relative to the ceptibility to false memories that develop through other, control events) lends evidence toward the formation guided imagination. However, it is useful to turn to the of a false memory for that event, and this effect has been small literature that has identified individual difference pre- termed imagination inflation (Garry et al., 1996). This dictors of false autobiographical memories across other phenomenon is often further supported by the partici- experimental manipulation types. Most of these individual pants’ report of a belief or an actual memory that the differences are concerned with traits that are captured by event occurred (see Bernstein et al., 2005a, 2005b for disengagement from reality. For example, one study studies that utilise memory-or-belief questionnaires). demonstrated that participants were more likely to falsely Source misattributions may help explain imagination remember a childhood event (ostensibly described by inflation. Whereas the misinformation effect is due to exter- their parents) if they were more prone to dissociation as nal source attributions (to the original event instead of to well as if they scored low on extraversion but were inter- the post-event narrative or questioning), imagination viewed by an experimenter who scored high on extraver- inflation is believed to be due to internal misattributions sion (Porter, Birt, Yuille, & Lehman, 2000). Another study (to one’s autobiography instead of to one’s imagination; utilising a similar technique also found a positive associ- Johnson et al., 1993). Another proposed explanation for ation with proneness to dissociation as well as fantasy pro- imagination inflation comes from the familiarity misattribu- neness (Hyman & Billings, 1998). In a study that employed tion hypothesis (Jacoby, Kelley, & Dywan, 1989), which guided visualisation, however, dissociation was not related explains that when an event is imagined, the imagination and participants’ self-reported extraversion, not introver- exercise renders the event more familiar and cognitively sion, was positively correlated with false memories available, which in turn results in false recollection. (Paddock et al., 2000). In yet another study in which partici- Researchers have demonstrated, however, that imagin- pants kept a diary of events and encountered an unex- ation inflation is more likely to occur when participants pected memory test six months later, memory intrusions 966 R. M. NICHOLS AND E. F. LOFTUS were found to be related to fantasy proneness but not dis- The methodological and theoretical considerations of sociation, absorption, or suggestibility (Horselenberg, each type of false memory combine to form a good launch- Merckelbach, van Breukelen, & Wessel, 2004). Absorption, ing point from which to consider whether performance in a construct closely related to hypnotisability and which one paradigm can itself predict performance in another. has been described as “a disposition for having episodes That is, to the extent that one falsely recognises the critical of ‘total’ attention that fully engage one’s representational lure in the DRM, will this predict his or her proneness to (i.e., perceptual, enactive, imaginative, and ideational) falsely recall an event from the past that never actually resources” (Tellegen & Atkinson, 1974), has been shown happened? Can one’s propensity to remember a critical to predict false memories related to the OJ Simpson trial event during an imagination inflation exercise predict the in a study of more naturally occurring memory intrusions overall likelihood of endorsing post-event information in (Platt et al., 1998). a typical misinformation experiment? Does the extent to As far as guided imagination is concerned specifically, which one remembers a misleading detail about an only a few studies have identified individual difference pre- event inform the likelihood of developing false memories dictors of developing a false memory. Working memory in a word-list task? has been shown to be negatively correlated with false memory for imagined events in the lab (Peters, Smeets, Empirical evidence Giesbrecht, Jelicic, & Merckelbach, 2007) and imagined A few studies have explored possible correlations between childhood events (Calvillo, Vasquez, & Pesavento, 2018). false memories arising from different paradigms. One study Proneness to dissociation has correlated positively with found that the 32% of children in their sample who devel- inflation (Heaps & Nash, 1999) though some studies have oped a false memory after an interview about a fictitious failed to find such a relationship (Horselenberg et al., event also displayed higher DRM rates compared to chil- 2000; Mazzoni & Memon, 2003). Hypnotic suggestibility dren who did not develop such a false memory (Otgaar, has also been shown as a predictor of inflation (Heaps & Verschuere, Meijer, & van Oorsouw, 2012). Among studies Nash, 1999), as has the ability to visualise or generate of adults, one reports a small positive correlation of r mental imagery (Horselenberg et al., 2000), though this = .12 between endorsement of misinformation in a stan- relationship has not held in other studies (Heaps & Nash, dard task and false memories in the DRM (Zhu, Chen, 1999; Mazzoni & Memon, 2003). It is clear from the litera- Loftus, Lin, & Dong, 2013). However, other studies have ture review of imagination inflation and of autobiographi- failed to find a relationship between misinformation and cal false memories more generally that there is mixed DRM false memories (Calvillo & Parong, 2016; Monds, evidence for many potential individual difference predic- Paterson, & Kemp, 2017; Ost et al., 2013), between DRM tors. This is especially true for those traits that cluster and false event suggestion (Otgaar & Candel, 2011), or around disengagement from conscious awareness such between imagination inflations for childhood events and as dissociation, absorption, hypnotisability, and fantasy for simple actions performed in the laboratory (Calvillo proneness. et al., 2018). Only two studies to date have examined cor- relations between false memories in misinformation, DRM, and false event suggestion manipulations, and Relatedness of paradigms both failed to find any correlations among them (Bernstein, Theoretical foundations Scoboria, Desjarlais, & Soucie, 2018; Patihis, Frenda, & The three paradigms reviewed thus far all reveal conditions Loftus, 2018). Overall, the research describes largely null under which memory for events can be manipulated and relationships between false memories of different para- distorted. Their theoretical underpinnings, however, vary. digms, but none so far have explored false memories The misinformation effect can be accounted for by specifically in the misinformation, DRM, and imagination source monitoring theory, which contends that people inflation paradigms all within subjects. This study aims to inaccurately remember misleading information as having address this need. occurred during the actual event and not from some post-event source. The DRM can be explained by a combi- A note on special populations nation of monitoring and activation theories in which There is some evidence that people who can recount extre- people misremember studying the critical item during mely anomalous or altogether impossible events are more encoding because it was cognitively activated instead of likely to commit memory errors in classic false memory actually presented and encoded. Finally, source monitoring experiments. The bulk of this evidence comes from confusions apply to a typical imagination exercise as well, research conducted with the DRM paradigm. One study though the confusion lies from internal, rather than exter- found that participants who self-reported at least one nal, sources. From a FTT perspective, gist traces support space alien abduction both falsely recognised and falsely false memory in both DRM and misinformation tasks, recalled more critical lures than a sample of non-abductee while verbatim traces support false memory in the misin- controls (Clancy et al., 2002). Another study similarly found formation paradigm (Reyna, Corbin, Weldon, & Brainerd, that people who report having past-lives (which is, under 2016). careful consideration, seemingly even less plausible than MEMORY 967 an alien abduction) also displayed heightened false recog- have been previously reported in the literature, (b) identify nition and recall rates in the DRM than controls who only new individual difference predictors of false memory for reported having one, current, on-going life (Meyersburg, one or more particular paradigms, (c) investigate which Bogdan, Gallo, & McNally, 2009). of these measures, if any, seem to commonly predict There has been some lingering debate regarding the false memories in all paradigms, and (d) use individual existence of memory repression (see Patihis, Ho, Tingen, difference correlates to refine proposed mechanisms for Lilienfeld, & Loftus, 2014; Brewin & Andrews, 2014; false memory development. Patihis, Lilienfeld, Ho, & Loftus, 2014), but it is of relevance to the discussion of special populations to describe research based on samples who report these types of Method these experiences. One study found that women who Participants reported recovering repressed memories of childhood sexual abuse (CSA) exhibited higher DRM rates than Undergraduate students from a large public university in women who never reported having forgotten the abuse California (N = 373; NFemale = 277, NMale = 89, NUnreported = (Clancy et al., 2000). The same pattern was found for a 7; Mage = 20.8, Asian 42.6%, Latino 25.1%, Caucasian/ group of women who reported repressing their CSA experi- White 15.3%, Middle Eastern 4.6%, African-American/ fi ences relative to women reporting either no abuse or no Black 1.6%, Hawaiian or Other Paci c Islander 0.5%, U.S. lapse in memory for their abuse (Geraerts et al., 2005). Fur- Indian 0.0%, Multiple ethnicities 9.0%, Other 1.1%) partici- thermore, a sample of women who reported recovering pated in the study for course credit. Of these participants, memories of traumatic childhood sexual abuse during psy- n = 297 completed the misinformation task, n = 283 com- chotherapy also exhibited higher false memory rates in the pleted the DRM task, n = 244 completed imagination fl ff DRM that women who experienced childhood abuse but in ation, and n = 367 completed individual di erence who were able to recall the event after simple relevant measures. Each of these tasks is described in further cues (Geraerts et al., 2009). detail below. The overarching idea of these studies is that those who are prone to developing a false memory from their past Materials due to an imagination, suggestion, or something else may also be susceptible to memory distortion in a labora- False memory paradigms tory setting. It is difficult to conceive a precise mechanism Misinformation. Participants completed a three-stage mis- that explains these patterns, in part because these studies information task adapted from Takarangi, Parker, and Garry are correlational, but also because false memories in the (2006) using Qualtrics software. In the first stage, partici- DRM and false autobiographical memories are the result pants viewed a silent 6 min 28 s video of a double-crossing of different manipulations and are themselves qualitatively electrician who steals personal property while on the job in dissimilar. It is perhaps the case, however, that the related- a client’s home. At the conclusion of the video, participants ness of these memories is most evident for people who lay responded to trivia statements for 12 min in order to at one extreme of the proneness spectrum for one or more prevent mental rehearsal of the events in the video. paradigms. It may also be true that they capture extreme Participants then read a narrative account of the events ends of trait distributions that are correlates of false mem- witnessed in the video. Each sentence in the narrative ories across different contexts. appeared one at a time, and participants were able to advance to the next sentence at his or her own pace. The narrative contained eight critical items from the video: Overview of the research bed, picture, van, time, mug, magazine, drink, and cap.In The first of two primary goals was to investigate the the narrative, four of the critical items were presented as relationships between memory performance in three misinformation items, and four were presented as control false memory paradigms administered within subjects: items. For example, Eric drinks from a can of Coke in the misinformation, DRM, and imagination inflation. Though video. In the narrative, the coke item either presented as the use of three paradigms is a methodological and logistic misinformation (“Eric helped himself to a can of Pepsi challenge, the benefit of this research design is that any from the fridge”) or with consistent information (“Eric pattern of results will facilitate understanding of the com- helped himself to a can of soft drink from the fridge”) monalities of false memories that are the result of that served for comparison. Participants were randomly different methodologies, that are themselves qualitatively assigned to one of two narrative conditions: in Condition unique, and that may be explained by different theoretical 1, four items in the narrative were misinformation items, mechanisms. and the other four items were controls. In Condition 2, Secondly, this study investigates a number of individual the reverse was true: misinformation items from Condition differences as predictors of false memories for each of the 1 were control items and control items from Condition 1 paradigms. The purpose of including these measures is to were misinformation items. After participants read the nar- (a) replicate patterns in a normal college-aged sample that rative, they completed basic addition and subtraction math 968 R. M. NICHOLS AND E. F. LOFTUS problems for five minutes again intended to prevent performed them for 15 s (18 statements), imagined per- rehearsal of items in the narrative. forming them for 15 s (18 statements), or only heard the The third and final stage of the task involved a recog- action and then completed math problems for 15 s (to nition test. Participants completed a 20-item, two-alterna- prevent mental rehearsal until the next statement was pre- tive forced-choice memory test about events in the sented; 18 statements). Half of the action statements original film of Eric the Electrician. For each item, partici- included an object (e.g., flatten the clay) and nine were non- pants also indicated whether they remembered their object action statements (e.g., nod in agreement). answers from the video, the narrative, both, or neither. Of After a one-week retention interval, participants the 20 items, four concerned events that contained mis- advanced through 44 action statements, presented one leading details in the narrative, four probed about events at a time on individual computers, for which they were that contained consistent information in the narrative, instructed to take a few moments to close their eyes and and the other twelve questions were fillers in which partici- imagine performing the action. After each imagination, pants addressed items they both viewed in the video and participants provided a vividness rating for what they read about in the narrative. For the eight critical questions, had imagined to help ensure compliance with the imagin- the answer alternatives consisted of the correct item (such ation instruction. After the imagination exercise, partici- as Coke) or the misinformation item (such as Pepsi). pants were reminded of their appointment for Session 3 Whether the incorrect alternative for each of these was a and dismissed. misinformation item or a foil depended on the condition Participants returned for the third and final study session the participant was in during the narrative stage. exactly one week after Session 2. This session began with the DRM exercise. Immediately following the DRM recog- DRM. Participants viewed the DRM list-learning task in a nition test, participants completed the final portion of the Microsoft PowerPoint slideshow also embedded into Qual- imagination inflation paradigm: a recognition test for the trics software. Before viewing the presentation, partici- action statements presented during Session 1. Each test pants were informed that they would see a series of item was followed by a confidence judgment, similar to words and that they should try to remember them to the the ones made in the DRM, from 1 (Not Confident) to 5 best of their ability. Immediately following this instruction, (Very Confident). Upon completion of Session 3, participants participants viewed 15 DRM word lists adapted from were thanked for their time, fully debriefed on the purpose Stadler et al., 1999 (anger, chair, doctor, fruit, king, moun- of the study, and awarded participation credit. tain, music, needle, rough, sleep, smoke, sweet, thief, trash, and window). Each word was presented for 1 s followed Individual differences by a 3 s pause between lists. At the conclusion of the Demographics. Participants filled out a basic demo- DRM presentation, participants completed the imagination graphics questionnaire that included questions about inflation test (see the following section), which also pro- their gender, age, ethnicity, number of years speaking vided a retention interval filler. Following that test, partici- English, current GPA, and SAT scores. pants completed a 120-item DRM recognition test. The recognition test contained 15 critical items (one from Trait personality. Personality dimensions of openness, each presented list), 45 studied items (taken from serial conscientiousness, extraversion, agreeableness, neuroti- positions 1, 8, and 10 of each list), and 60 new, unrelated cism were measured using the 44-item Big Five Inventory distractor items (also drawn from critical items and items (John, Donahue, & Kentle, 1991; John, Naumann, & Soto, in serial positions 1, 8, and 10, of lists not used in the pres- 2008). Participants rated their agreement with each item entation). For each item in the recognition test, participants on a 5-point Likert scale (1 = Disagree Strongly; 5 = Agree judged whether they had seen it (“OLD”) or had not seen it Strongly). (“NEW”) previously in the slideshow and provided a confi- dence judgment for their decision using a scale anchored Cognitive ability and style. Overall cognitive ability was at 1 (Not Confident) to 5 (Very Confident). For each item gauged with the 150-item Over-claiming Questionnaire judged “OLD”, participants also indicated whether they (OCQ; Paulhus & Harms, 2004; Paulhus, Harms, Bruce, & had a specific memory for having seen the word (“REMEM- Lysy, 2003), which asks participants to rate their familiarity BER”) or if they knew it had been presented but did not with a number of historical, scientific, and cultural items experience an actual recollection of the word or could on a 5-point Likert scale (0 = never heard of it; 4 = very fam- remember any details about it (“KNOW”). These indices iliar). To measure cognitive style, participants completed were used to distinguish memories from beliefs, both the 18-item Need for Cognition (NFC) subscale and respectively. 12-item Faith in Intuition (FI) subscale of the Rational-Experi- ential Inventory (REI; Epstein et al., 1996). For both scales, Imagination inflation. The imagination inflation task, participants indicated how characteristic each statement adapted from Goff and Roediger (1998), took place over is of them using a Likert scale (1 = extremely uncharacteris- three in-lab sessions. In Session 1, participants heard 54 tic; 7 = extremely characteristic). Participants also com- action short statements (e.g., push the toy car) and either pleted the Cognitive Reflection Task (CRT; Frederick, 2005), MEMORY 969 a three-item measure designed to gauge a subject’s willing- Table 1. Study procedures. ness and ability to suppress an intuitive but incorrect Session 1 answer to a problem and arrive at a correct answer. Study information sheet Finally, participants completed the Vividness of Visual Individual difference measures Hear, imagine, or perform imagination inflation action statements Imagery Questionnaire (VVIQ; Cui, Jeter, Yang, Montague, Session 2 (one week after Session 1) & Eagleman, 2007; Marks, 1973), which measures how Misinformation video vividly one is able to produce mental images. Participants Filler task visualised a number of statements and indicated the vivid- Misinformation narrative ness of each visual image using a Likert scale (1 = no image, Filler task Misinformation test you are only thinking of the object; 5 = image is perfectly Imagine imagination inflation action statements clear and as vivid as normal vision). Session 3 (one week after Session 2) DRM slideshow Disengagement from reality. Participants completed the DRM recognition test Dissociative Experiences Scale (DES C; Wright & Loftus, Imagination inflation recognition test fi 1999), a measure designed to investigate proneness to dis- Debrie ng Notes: The misinformation is paradigm was contained entirely in Session 2, sociation, by rating the frequency of such experiences on an the DRM paradigm was contained entirely in Session 3, and the imagin- 11-point scale. Absorption was measured with the Tellegen ation inflation paradigm spanned all three sessions. Subjects completed Absorption Scale (TAS; Tellegen & Atkinson, 1974), for which all individual difference questionnaires in Session 1. participants rated the frequency of absorptive items on a 4- point scale (1 = never; 4 = always). Finally, participants also Session 2 occurred exactly one week after Session 1, and completed a measure of fantasy proneness, the Creative all elements of Session 2 took place on the participants’ indi- Experiences Questionnaire (CEQ; Merckelbach, Horselen- vidual computers. When participants first arrived, they com- berg, & Muris, 2001). For this scale, participants indicated pleted the entire misinformation paradigm on the whether or not they experienced as a child, or tend to computer: they viewed the original event, completed the experience as an adult, events that that involve fantasising. trivia-question filler task, read the narrative containing mis- information, worked on the math-problem filler task, and Religiosity. Participants’ internal and external religiosity then completed the forced-choice recognition test about was measured using the 20-item Age Universal Religious original event details. Upon completion, participants Orientation Scale (AUROS; Gorsuch & Venable, 1983). engaged in the second part of the imagination inflation pro- cedure. During this time, 144 action statements appeared to Anomalous experiences. Self-reported paranormal and/or participants one at a time on their computer screens. Partici- anomalous experiences were investigated with the Anoma- pants were instructed that for each statement, they should lous/Paranormal Experience Subscale of the Anomalous take a few moments to close their eyes and imagine per- Experiences Inventory (AEI; Gallagher, Kumar, & Pekala, 1994). forming each action. After each imagination, participants provided a vividness rating for what was imagined to help Procedure ensure that they were imagining the actions. After partici- pants completed these imaginations and their vividness The order of events for all three sessions appears in Table 1. ratings on their own time, they were reminded of their Study procedures occurred over three separate in-lab ses- appointment for Session 3 and dismissed. sions in which up to five participants participated at a time. Participants returned for their third and final session When participants arrived for Session 1, they first read an exactly one week after Session 2. This session began with IRB-approved study information sheet and consented to the DRM slideshow and recognition test. Immediately fol- participate in all subsequent research activities. Partici- lowing the test, participants completed the final portion pants were also informed that they could withdraw from of the imagination inflation paradigm: a recognition test participation at any time without receiving any penalty. for the action statements presented during Session 1. All After consenting to participate, participants completed components of Session 3 were also completed on individ- the demographics questionnaire followed by all other indi- ual computers. At the completion of Session 3, participants vidual difference measures presented in a randomised were thanked for their time, fully debriefed on the purpose order. At the end of the questionnaires, participants com- of the study, and awarded participation credit through the pleted an attention check (adapted from Oppenheimer, university’s human subject pool. Meyvis, & Davidenko, 2009). All procedures up to this point were administered on individual computers. They then engaged in the first part of the imagination inflation Results exercise in which they either performed, imagined, or Attention check merely heard the 54 action statements. After the partici- pants heard all 54 action statements, they were reminded Of the 367 participants who encountered the attention of their two future in-lab sessions and dismissed. check, 21 did not pass. However, all study procedures 970 R. M. NICHOLS AND E. F. LOFTUS took place in the lab under the supervision of a research The Deese-Roediger-McDermott (DRM) paradigm assistant, and their recognition memory in all three para- Recognition rates. The recognition rate for studied items digms did not differ significantly from those who passed was fairly high (M = .68, SD = .20) and that of distractor the check. Therefore, their data were included in all items fairly low (M = .30, SD = .18). Critical lures were analyses. falsely recognised at a fairly high rate as well (M = .63, SD = .15). Tests of within-subjects contrasts revealed a signifi- cant difference between recognition rates of critical and False memories studied items, though the size of the effect is small, F(1, h2 = . The nature of this research requires a sufficient supply of 281) = 17.33, p < .001, p 06. There was a strong, posi- memory distortions to be able to explore the relationships tive correlation between endorsement rates of studied between them. To preview the findings, all three memory items and critical lures, r = .60, p < .001, illuminating the manipulations reliably produced false memories. Further- associative nature of memory in this paradigm. However, fi ff more, all participants developed false memories of some there is a large and statistically signi cant di erence kind. Of the 297 participants who completed the misinfor- between the mean recognition rates of critical and mation paradigm, only eight failed to falsely endorse any studied items combined and the recognition rate of dis- h2 = . misinformation items. Of the 283 participants who com- tractor items, F(1, 281) = 1219.77, p < .001, p 81. Of pleted the DRM, only two others did not falsely recognise the 15 DRM lists utilised, both the king and mountain lists any critical lures. Of the 244 participants who completed produced the lowest endorsement rates of their critical the imagination inflation exercise, only one other failed lures (M = .51, SD = .50) while both the sweet and window to judge any unheard but later imagined statements as lists produced the highest endorsement rate of their critical heard. lures (M = .79, SD = .41). Internal consistency of recognition rates was α = .67 for critical items (15 items), α = .82 for studied items was (45 items), and α = .91 for distractor The misinformation paradigm items (60 items). fi Memory and con dence judgments. Participants were When participants indicated an item as studied at test, more accurate for control items (M = .77, SD = .21) than they were also asked to indicate whether they had an for items for which they were misled (M = .40, SD = .27), t actual memory for having seen an item or mere belief ’ (296) = 19.18, p < .001, Cohen s d = 1.55. Participants were that they had seen the item (but without any sense of a fi equally con dent, however, when making judgments for memory). Robust false DRM memories are defined as control items (M = 4.25, SD = .66) and misled items (M = those false recollections of critical lures that are 4.19, SD = .74), t(296) = 1.26, p = .21, d = .14. For every one accompanied by a memory judgment. Of all critical lures of the eight critical items, participants were more accurate presented at test, a sizeable minority of them were recog- when it appeared for them as a control item than when it nised robustly (M = .20, SD = .16), and this proportion is not appeared for them as a misled item. A paired t-test statistically different from the proportion of studied items revealed that overall, items were better remembered as robustly recognised (M = .20, SD = .12; t(277) = 1.06, p controls than as misled items, t(7) = 7.36, p < .001, d= = .29, d = .07). These data support the notion that judg- 2.55. Because there are only four misinformation items ments about critical lures are not simply based on criterion and four control items that are counterbalanced for each shifts but rather indeed represent actual memory distor- half of the participants, reliability estimates for these tion phenomena. items were not calculated. Confidence judgments. Participants were equally Source monitoring judgments. For each test item, partici- confident for their judgments about critical (M = 3.23, SD pants indicated whether they remembered their answers = .69) studied (M = 3.23, SD = .66), and distractor items (M from the video, the narrative, both, or neither. False mem- = 3.23, SD = .65), Huynh-Feldt F(1.63,453.73) = .05, p = .92. ories of critical items combined with source misattributions Confidence within subjects for these three items is also to either the video or to both the video and the narrative highly correlated, adjusted multiple r2 = .91. Adjusted mul- constitute robust false memories. The proportions of tiple r2 statistics were calculated using the following robust false memories for misleading details in the misin- formula: formation condition (M = .42, SD = .25) is significantly 




















2 + 2 − higher than the proportion of the same source misattribu- rxz ryz 2rxzryzrxy r . = tions for falsely-endorsed misleading details in the control z xy − 2 1 rxy condition (M = .15, SD = .16), t(292) = 15.65, p < .001, d = 1.30. When robust false memories are calculated as pro- portions of all falsely remembered items, rather than out The imagination inflation paradigm of all test items, the proportions are approximately equal Memory for having heard Session 1 statements. To between critical and control items t(186) = .72, p = .47, d understand how imagination during Session 2 affected = .07. memory for action statements presented in Session 1, MEMORY 971

Figure 1. Mean rates of Heard judgments for imagination inflation items by Session 1 presentation type and Session 2 imagination type. Error bars represent 95% within-subjects confidence intervals.

’ h2 = . participants memories were tested during Session 453.48) = 948.53, p <.001, p 80. This relationship is 3. Specifically, at test, participants were presented with a further qualified by a significant interaction, Huynh-Feldt h2 = . ff number of action statements (including those presented F(1.87, 453.48) = 948.53, p < .001, p 80. The e ect of in Session 1 as well as some new, never-presented state- imagination, or the magnitude of the difference in memory ments) and indicated whether they remembered hearing for events between not imagining, imagining once, and ima- the statement during Session 1. They also provided a confi- gining five times, was greatest for unheard statements. dence judgment for their answer, again using a scale Moreover, a one-way repeated measures ANOVA revealed anchored at 1 (Not Confident) to 5 (Very Confident). large differences in memory between non-imagined h2 = . Furthermore, if they did remember hearing the action Session 1 statements, F(3, 729) = 456.67, p < .001, p 65. statement, they were asked to indicate whether they per- Those differences are smaller for those statements imagined h2 = . ff formed, imagined, or only heard the action. once, F(3, 729) = 257.48, p < .001, p 51, though the di er- In order to analyze the memory data, each block of state- ence still represents a large effect. For statements that ments was first coded according to what the participant was were imagined five times, the effect is smaller yet, Huynh- h2 = . instructed to do with it in Session 1 as well as the number of Feldt F(2.762, 671.19) = 64.47, p <.001, p 22. times that they imagined it in Session 2. This results in 12 cat- egories of items at test; mean recognition rates for these Confidence judgments for having heard Session 1 state- items appear in Figure 1. A 4 (Session 1 Item Type) × 3 ments. Recall that participants provided a confidence (Number of Imaginations in Session 2) repeated-measures rating for their judgments of whether or not they heard a ANOVA was then performed to gauge how often participants particular statement during Session 1. A two-way repeated remembered having heard these items in Session 3. For measures ANOVA was performed on confidence judg- the following analyses, wherever assumptions of sphericity ments, and this test reveals a significant effect of Session are violated, Huynh-Feldt F statistics are reported. h2 = . 1 Item Type, F(3, 732) = 4.90, p = .002, p 02. This There was a main effect of Session 1 Item Type: partici- effect is small, though post-hoc analyses (Helmert con- pants made correct Heard judgments most often to items trasts) indicate that confidence for unheard statements they had performed in Session 1 (M = .87, SD = .13). Their was significantly lower than for the other types of state- rates of correct recognition were lower if they had imagined h2 = . ments combined, F(1, 244) = 13.17, p < .001, p 05. the item in Session 1 (M = .67, SD = .16) and lower still if they However, there was no effect of Session 2 imagination con- had merely heard the item in Session 1 (M = .54, SD = .11). dition (p = .22) or interaction of item type by imagination Unheard items were recognised as heard least often (p = .84) on confidence ratings. In other words, across all (M = .44, SD = .16), Huynh-Feldt F(2.83, 688.45) = 575.11, Session 1 item types, memory judgments for actions h2 = . ff p < .001, p 70. Furthermore, there was a main e ect of were made with equal confidence regardless of whether imagination; statements that were not imagined at all in they were not imagined, imagined once, or imagined five Session 2 were recognised less often (M = .41, SD = .14) times. Furthermore, the range of confidence judgments than statements that were imagined once (M = .65, SD for all 12 item types (all 12 cell means) is small; these = .15) or five times (M = .83, SD = .14), Huynh-Feldt F(1.87, scores range from 3.73 to 3.87 on a 5-point scale. 972 R. M. NICHOLS AND E. F. LOFTUS

Figure 2. Mean rates of Performed judgments for imagination inflation items by Session 1 presentation type and Session 2 imagination type. Error bars rep- resent 95% within-subjects confidence intervals.

Memory for having performed Session 1 statements. misinformation items recognised at test after having Participants who responded during Session 3 that they encountered those items in the post-event narrative and had indeed heard a statement during Session 1 were also the proportion of these items that were robustly recog- asked to indicate whether they remembered performing, nised (i.e., that were recognised and that were imagining, or merely listening to that statement during accompanied by a judgment of having seen the item in Session 1. Items judged as performed are depicted in either the video or the video and the narrative). In the Figure 2, and the pattern of data is similar to the proportion DRM, these measures include the proportion of critical of remembered statements. Another 4 (Session 1 Item lures falsely recognised and the proportion of critical Type) × 3 (Number of Imaginations in Session 2) lures that are robustly recognised (i.e., that are repeated-measures ANOVA revealed a main effect of accompanied by a Memory judgment). Correlations Session 1 Item Type on the proportion of remembered between these measures appear in Table 2. There is a sig- items deemed performed, Huynh-Feldt F(2.81, 684.79) = nificant positive correlation between the endorsement rate h2 = . ff 65.21, p < .001, p 21, and a main e ect of imagination, of critical lures in the DRM and both the proportion of mis- h2 = . Huynh-Feldt F(1.85, 448.22) = 90.25, p < .001, p 27. information items endorsed, r = .14, p = .03, as well as the Again, a significant interaction was also found, Huynh- proportion of misinformation items robustly endorsed, r h2 = . Feldt F(5.60, 1360.36) = 6.03, p < .001, p 02. = .17, p = .01. There are only trend-level relationships, It is worth noting that with zero imaginings, participants however, between the proportion of robustly recognised reported actually having performed statements that they critical lures and both the proportion of misinformation did not even hear 3.01% of the time. This percentage items endorsed, r = .11, p = .09, and the proportion of mis- increases to 9.16% of the time for unheard statements that information items robustly endorsed, r = .11, p = .08. were imagined once and increases even higher to 15.90% of the time for unheard statements that were imagined fl five times. This difference is statistically significant, Huynh- The misinformation and imagination in ation h2 = . paradigms Feldt F(2.04, 410.44) = 69.94, p < .001, p 22. These mem- ories are perhaps most impressive in that participants claim Both false memory measures in the misinformation para- to have heard and performed actions – like stacking checkers digm were correlated with multiple false memory or faking sneezes – that they never even heard. Table 2. Pearson R correlations between false memory measures in the misinformation and DRM paradigms. Inter-paradigm relationships Misinformation measures In order to understand one’s propensity to develop false Proportion of Proportion of misinformation items misinformation items memories across multiple tasks, correlations were analysed DRM measures recognised robustly recognised between all three false memory tasks. Each of these Proportion of critical 0.14* 0.17** relationships is described separately below. lures recognised Proportion of critical 0.11† 0.11† lures robustly The misinformation and DRM paradigms recognised Both tasks produced two measures of false memory. In the Note: Correlations in bold are statistically significant at p < .05 or smaller. † misinformation task, these measures are the proportion of p < .10; *p < .05; **p < .01. MEMORY 973

Table 3. Pearson R correlations between false memory measures in the misinformation and imagination inflation paradigms. Misinformation measures Proportion of misinformation Proportion of misinformation Imagination inflation measures items recognised items robustly recognised Proportion of Heard judgments for unheard items in Session 1 Items imagined once in Session 2 −0.01 0.00 Items imagined five times Session 2 0.14* 0.11† Items imagined at all in Session 2 0.08 0.07 Proportion of Performed judgments for non-performed items in Session 1 Items unheard in Session 1 Items imagined once in Session 2 −0.04 −0.12† Items imagined five times Session 2 0.00 −0.02 Items imagined at all in Session 2 −0.02 −0.09 Items listened-to in Session 1 Items imagined once in Session 2 −0.01 0.04 Items imagined five times Session 2 0.04 0.15* Items imagined at all in Session 2 0.02 0.13* Items imagined in Session 1 Items imagined once in Session 2 0.08 0.08 Items imagined five times Session 2 0.07 0.04 Items imagined at all in Session 2 0.10 0.08 Note: Correlations in bold are statistically significant at p < .05 or smaller. †p < .10; *p < .05.

measures in the imagination inflation paradigm. Namely, correlations appear in Table 4. The proportion of critical false memories of interest include Heard judgments to lures falsely recognised correlates significantly with the statements unheard in Session 1 but imagined in Session proportion of Performed judgments to items listened to 2aswellasPerformed judgments to statements that in Session 1 and imagined five times in Session 2, r = .17, were unheard, merely listened to, or imagined in Session p = .01. This relationship holds when the imagination vari- 1 and later imagined in Session 2. The results of the corre- ables includes listened-to items that were imagined at all, lations of these measures to misinformation false memory r = .16, p = .02. measures appear in Table 3. It is worth noting that corre- When examining the relationships between imagin- lations involving imagination inflation items that were ima- ation inflation false memories and robustly recognised criti- gined “at all” represent a collapse across items imagined cal lures, additional significantly positive associations once and items imaged five times. emerge. For example, the proportion of robustly recog- There is a significant positive correlation between the nised critical lures is positively correlated with the pro- proportion of misinformation items recognised and the portion of Heard judgments for unheard statements in proportion of Heard judgments for unheard items in Session 1 that are imagined once, r = .13, p = .04, imagined Session 1 that were imagined five times in Session 2, r five times, r = .13, p = .04, or imagined at all r = .17, p < .01. = .14, p = .03. When the misinformation variable in this Furthermore, robustly recognised lures are correlated with relationship is the proportion of robustly recognised the proportion of Performed judgments to unheard items in items, the correlation remains positive but only at a trend Session 1 and imagined once in Session 2, r = .17, p < .01. level of significance, r = .11, p = .08. There is also significant, For items that are merely listened-to in Session 1, robustly positive correlation between the proportion of misinforma- recognised critical lures are associated with Performed tion items robustly recognised and the proportion of Per- judgments to these items when they are imagined once, formed judgments to items listened to in Session 1 and r = .17, p = .01, or imagined at all, r = .19, p < .01; this imagined five times in Session 2, r = .15, p = .02. This relationship approaches significance when these items relationship holds when the imagination variable rep- are imagined five times, r = .12, p = .05. Finally, for items resents listened-to statements that are imagined either that are imagined in Session 1, robustly recognised critical once or five times, r = .13, p = .04. However, there is also a lures are associated with Performed judgments to these trend-level negative relationship between the proportion items when they are imagined once, r = .14, p = .03, and of misinformation items robustly recognised and the pro- imagined at all, r = .13, p = .04. portion of Performed judgments to unheard items in Session 1 that were imagined once in Session 2, r = −.12, Confidence p = .06. Participants rated their confidence on a scale of 1 (Not Confident) to 5 (Very Confident) for their DRM recognition The DRM and imagination inflation paradigms judgments, judgments for misinformation test items and Both false memory measures in the DRM paradigm were judgments of whether a particular imagination inflation correlated with the same false memory measures in the item was heard or not (note that confidence was not imagination inflation paradigm described above. These assessed for judgments about whether an item deemed 974 R. M. NICHOLS AND E. F. LOFTUS

Table 4. Pearson R correlations between false memory measures in the DRM and imagination inflation paradigms. DRM measures Proportion of critical Proportion of critical lures Imagination inflation measures lures recognised robustly recognised Proportion of Heard judgments for unheard items in Session 1 Items imagined once in Session 2 0.04 0.13* Items imagined five times Session 2 0.10 0.13* Items imagined at all in Session 2 0.09 0.17** Proportion of Performed judgments for non-performed items in Session 1 Items unheard in Session 1 Items imagined once in Session 2 0.01 0.17** Items imagined five times Session 2 −0.02 0.00 Items imagined at all in Session 2 −0.01 0.09 Items listened-to in Session 1 Items imagined once in Session 2 0.04 0.17** Items imagined five times Session 2 0.17** 0.12† Items imagined at all in Session 2 0.16* 0.19** Items imagined in Session 1 Items imagined once in Session 2 0.03 0.14* Items imagined five times Session 2 0.05 0.05 Items imagined at all in Session 2 0.05 0.13* Note: Correlations in bold are statistically significant at p < .05 or smaller. †p < .10; *p < .05; **p < .01. heard in Session 1 was performed, imagined, or merely lis- the misinformation task (r = .13, p = .03), but no other indi- tened-to at that time). There is a trend-level positive corre- vidual difference variables predict overall accuracy, robust lation between confidence for misinformation items and accuracy, or confidence judgments for control items. confidence for critical lures, r = .11, p = .09, and a significant fi positive correlation between con dence for misinforma- The DRM paradigm fi tion items and con dence for unheard, imagined items, r Correlations were examined between the same individual = .19, p < .01. A strong positive correlation exists between difference variables and measures of memory in the DRM fi fi con dence for critical lures and con dence judgments task. Several more relationships emerge for the DRM than for unheard, imagined items, r = .59, p < .001. All three for the misinformation paradigm; these correlations appear fi con dence judgments are also positively correlated, in Table 7.Specifically, total false recognition of critical 2 adjusted multiple r = .15. lures is associated with the Cognitive Reflection Task (CRT) in terms of both the number of intuitive answers given in the task (r = .25, p < .001) and the number of correct Individual differences answers given to the task (r = −.20, p = .001). The number Table 5 contains descriptive statistics for the demographic of correct answers given to the CRT is also associated with and personality variables measured in this investigation. a decrease in robustly recognised critical lures (r = −.12, p These descriptives include the number of observations, = .04). A higher score on the Anomalous/Paranormal Experi- the minimum and maximum scores, the mean and stan- ence Subscale of the Anomalous Experiences Inventory (AEI), dard deviation, the skewness and kurtosis, and Cronbach’s on the other hand, is associated with increase in robustly alpha (where relevant) of each variable. The relationships recognised lures (r = .13, p = .04). In terms of confidence of these variables to memory phenomena are first dis- judgments for critical lures, only OCQ discrimination scores cussed within the context of each paradigm. are associated such judgments (r = −.23, p < .001). When examining veridical memory, overall recognition The misinformation paradigm of studied items was associated only with the sum of intui- Correlational analyses were conducted to examine which tive answers given to the CRT (r = .13, p = .03). Robustly relationships, if any, exist between demographic/personal- recognised studied items are correlated only with the Crea- fi ity variables and memory measures in the misinformation tive Experiences Questionnaire (r = .12, p = .04). Con dence paradigm. These Pearson R correlations appear in judgments for studied items, on the other hand, is associ- − Table 6. As this table shows, there are no significant individ- ated with OCQ discrimination scores (r = .21, p < .001), ual difference correlates with either overall or robust false absorption (r = .13, p = .03), and religiosity (r = .13, p = .03). memory. However, three personality scales are associated with confidence judgments for misinformation items: con- The imagination inflation paradigm scientiousness (r = .19, p < .001), Need for Cognition (r = .12, As was calculated for the DRM, correlations were computed p = .04), and dissociative experiences (r = −.16, p = .01). Dis- between measures of memory in the imagination inflation crimination scores of the Over-claiming Questionnaire are paradigm and individual difference variables of interest; positively associated with accuracy for control items in these correlations appear in Table 8. For the sake of MEMORY 975

Table 5. Descriptive statistics for demographic and personality measures. Measure N Min Max MSDSkewness Kurtosis Cronbach’s α Age 367 18.00 55.00 20.77 3.64 4.27 27.66 – Proportion of life speaking English 360 .10 1.00 0.90 0.19 −2.20 4.60 – Estimated GPA 342 .50 4.50 3.11 0.59 −0.55 0.55 – Big Five – Extraversion 365 1.00 6.00 3.37 0.82 −0.18 −0.20 0.86 Big Five – Agreeableness 365 2.13 6.00 3.88 0.59 −0.26 −0.03 0.71 Big Five – Conscientiousness 365 1.89 6.00 3.64 0.65 0.09 −0.09 0.77 Big Five – Neuroticism 365 1.00 6.00 3.10 0.83 −0.03 −0.15 0.84 Big Five – Openness 365 1.60 6.00 3.51 0.61 −0.06 0.66 0.76 OCQ – Discrimination 366 −2.52 3.37 0.00 1.00 0.18 0.00 – OCQ – Bias 366 −2.70 2.17 0.00 1.00 −0.52 0.16 – NFC 364 24.00 125.00 80.09 18.20 −0.08 0.04 0.86 FI 365 28.00 84.00 57.94 11.11 −0.23 −0.16 0.85 CRT – Sum of Intuitive Answers 364 0.00 3.00 2.05 1.02 −0.71 −0.71 – CRT – Sum of Correct Answers 364 0.00 3.00 0.59 0.89 1.38 0.84 – VVIQ 365 1.81 5.00 3.80 0.68 −0.40 −0.46 0.95 DES 365 0.00 100.00 29.94 16.89 0.62 0.51 0.94 TAS 363 2.00 51.00 17.12 8.30 0.72 0.65 0.87 CEQ 364 0.00 22.00 9.26 4.44 0.10 −0.41 0.76 AUROS 334 10.00 100.00 52.11 16.33 −0.36 −0.70 0.91 AEI 364 0.00 20.00 4.97 3.31 1.34 2.71 – parsimony, false memories are represented in these ana- second is Performed judgments to statements that were lyses through only two variables. The first is the proportion performed in Session 1 and not imagined in Session of Heard judgments to statements unheard in Session 1 2. Accurate memories operationalised this way are most and imagined at all in Session 2 (collapsed across state- akin to memory accuracy for control items in the misinfor- ments imagined once and statements imagined five mation paradigm. times). The second is the proportion of Performed judg- In terms of false memory, several individual difference ments to statements that were not performed in Session measures were correlated with both false Heard judgments 1 (collapsed across unheard, listened-to, and imagined and Performed judgments after imagination. These statements) and imagined at all in Session 2 (again col- includes trait agreeableness (Heard: r = −.14, p = .03; Per- lapsed across statements imagined once and statements formed: r = −.14, p = .03), OCQ Bias scores (Heard: r = −.14, imagined five times). True memories were assessed p = .03; Performed: r = −.18, p < .01), and dissociative experi- through two similarly constructed variables. The first is ences (Heard: r = .13, p < .05; Performed: r = .20, p < .01). Fur- Heard judgments to statements that were heard in thermore, false Performed judgments are also associated Session 1 (collapsed across listened-to, imagined, and per- with absorption (r = .15, p = .02) and anomalous experi- formed statements) and not imagined in Session 2. The ences (r = .13, p < .05). Confi dence judgments for

Table 6. Pearson R correlations between demographic and personality variables and misinformation variables. Misinformation items Control items Overall false Robust false Overall veridical Robust veridical Measure memory memory Confidence memory memory Confidence Age −0.03 0.01 0.03 0.07 −0.04 0.09 Proportion of life speaking English 0.01 0.03 0.05 0.09 −0.08 −0.06 Estimated GPA −0.08 0.04 0.09 0.00 −0.02 −0.02 Big Five – Extraversion 0.04 0.06 0.07 0.04 0.03 0.04 Big Five – Agreeableness −0.01 0.01 0.06 0.00 0.00 −0.05 Big Five – Conscientiousness 0.04 0.09 0.19*** −0.01 0.05 0.06 Big Five – Neuroticism 0.04 −0.03 −0.04 −0.07 −0.01 −0.04 Big Five – Openness 0.05 0.04 0.00 0.00 0.04 −0.04 OCQ – Discrimination −0.10 −0.03 0.05 0.13* −0.07 0.00 OCQ – Bias 0.01 0.08 0.05 0.07 −0.01 −0.01 NFC −0.10 −0.01 0.12* 0.00 0.04 0.04 FI 0.01 −0.01 0.03 0.01 −0.01 −0.01 CRT Intuitive Answers 0.06 0.06 0.04 0.00 0.03 0.04 CRT Correct Answers −0.07 −0.05 0.03 0.07 −0.08 0.00 VVIQ −0.02 0.06 0.02 −0.01 −0.02 0.06 DES 0.01 −0.06 −0.16** −0.09 0.03 −0.09 TAS −0.01 −0.06 −0.11 −0.08 0.01 −0.02 CEQ 0.07 0.03 −0.07 −0.08 0.06 −0.02 AUROS 0.10 0.07 0.03 −0.02 0.00 −0.05 AEI 0.07 −0.02 −0.05 −0.03 0.00 −0.06 Note: Correlations in bold are statistically significant at p < .05 or smaller. *p < .05; **p < .01; ***p < .001. 976 R. M. NICHOLS AND E. F. LOFTUS

Table 7. Pearson R correlations between demographic and personality variables and DRM variables. Critical lures Studied items Overall false Robust false Overall veridical Robust veridical Measure memory memory Confidence memory memory Confidence Age −0.06 0.07 −0.10 −0.05 −0.03 −0.08 Proportion of life speaking English 0.10 0.06 −0.09 0.01 0.01 −0.07 Estimated GPA −0.10 −0.03 −0.09 −0.10 −0.04 −0.12 Big Five – Extraversion −0.02 0.08 0.02 0.03 0.05 0.02 Big Five – Agreeableness −0.03 −0.08 0.10 −0.02 −0.03 0.09 Big Five – Conscientiousness −0.01 −0.01 0.07 0.02 −0.06 0.08 Big Five – Neuroticism 0.08 0.02 0.00 0.10 0.02 −0.03 Big Five – Openness −0.06 0.01 −0.01 −0.03 0.03 0.00 OCQ – Discrimination −0.05 −0.05 −0.23** −0.02 −0.07 −0.21*** OCQ – Bias 0.04 −0.02 0.02 0.01 −0.07 0.04 NFC −0.07 −0.04 −0.07 −0.05 −0.04 −0.09 FI −0.02 0.10 0.05 −0.01 0.11 0.08 CRT Intuitive Answers 0.25*** 0.09 0.01 0.13* 0.09 0.01 CRT Correct Answers −0.20** −0.12* −0.03 −0.06 −0.11 −0.04 VVIQ −0.02 0.10 0.09 −0.02 0.10 0.05 DES −0.06 −0.02 0.02 −0.02 −0.01 0.05 TAS −0.05 0.02 0.09 0.00 0.08 0.13* CEQ 0.08 0.02 0.03 0.09 0.12* 0.06 AUROS 0.06 0.11 0.11 0.05 0.07 0.13* AEI 0.01 0.13* 0.03 0.02 0.11 0.05 Note: Correlations in bold are statistically significant at p < .05 or smaller. *p < .05; **p < .01; ***p < .001. unheard, later imagined statements is associated with both Discussion OCQ discrimination (r = −.15, p = .02) and OCQ bias (r = .13, Inter-paradigm relationships p = .04). In the absence of any imagination, accurate Heard judg- We assessed overall and robust false memories in the mis- ments are associated with OCQ bias (r = −.21, p = .001) and information paradigm, overall and robust false memories in dissociative experiences (r =.14,p = .03). Accurate Performed the DRM, and false Heard and false Performed judgments to judgments are also positively associated with dissociative several items in the imagination inflation task. Small but experiences, (r =.13, p = .04). Confidence for items that statistically significant correlations emerged between were heard but not imagined is associated with estimated some false memory variables in each of the three compari- GPA (r = −.16, p = .02), OCQ discrimination (r = −.17, p < .01), sons. Every significant correlation is positive, but again, Faith in Intuition (r =.17,p < .01), religiosity (r =.19,p < .01), they are typically rather modest: Pearson R coefficients and anomalous experiences (r = .13, p <.05). range from r = .14 to r = .17 between false memories in

Table 8. Pearson R correlations between demographic and personality variables and imagination inflation variables. False memories True memories Measure Heard judgments Performed judgments Confidence Heard judgments Performed judgments Confidence Age 0.04 0.06 0.03 −0.08 0.00 −0.06 Proportion of life speaking English −0.05 −0.04 −0.02 0.01 −0.01 −0.08 Estimated GPA −0.01 −0.10 −0.09 −0.07 0.01 −0.16* Big Five – Extraversion −0.06 0.01 0.01 0.07 −0.03 0.08 Big Five – Agreeableness −0.14* −0.14* 0.08 −0.09 −0.04 0.07 Big Five – Conscientiousness −0.06 −0.01 0.09 −0.05 −0.11 0.05 Big Five – Neuroticism 0.06 0.04 −0.04 0.00 −0.01 0.04 Big Five – Openness −0.03 0.04 0.01 0.04 −0.03 0.03 OCQ – Discrimination 0.04 −0.05 −0.15* −0.02 −0.09 −0.17** OCQ – Bias −0.14* −0.18** 0.13* −0.21** −0.08 0.08 NFC 0.00 −0.05 −0.03 0.04 −0.10 −0.04 FI 0.00 0.04 0.08 −0.05 0.02 0.17** CRT Intuitive Answers 0.03 −0.08 0.09 −0.03 −0.05 0.05 CRT Correct Answers 0.04 0.00 −0.12 0.03 −0.01 −0.02 VVIQ 0.00 −0.03 0.04 −0.04 0.00 0.09 DES 0.13* 0.20** 0.02 0.14* 0.13* 0.03 TAS 0.06 0.15* 0.11 0.07 0.04 0.12 CEQ 0.03 0.08 0.11 0.05 0.09 0.11 AUROS −0.02 0.12 0.12 −0.01 0.00 0.19** AEI 0.12 0.13* 0.06 0.00 0.05 0.13* Note: Correlations in bold are statistically significant at p < .05 or smaller. *p < .05; **p < .01; ***p < .001. MEMORY 977 the misinformation paradigm and DRM (consistent with place entirely in Session 2. It could be the case that a Zhu et al., 2013), r = .13 to r = .15 between misinformation global sense of confidence during Session 3 affected feel- false memories and imagined false memories, and from r ings of certainty for memory judgments, and perhaps if = .13 to r = .19 between false memories in the DRM and the misinformation test was also given in Session 3, the cor- imagined false memories. relations between confidence judgments for misinforma- Concerning the misinformation and imagination tion items would be stronger. inflation tasks, correlations emerged between total false It is no surprise that these correlations are small for misinformation memories and false Heard judgments to some comparisons and altogether absent in many of the items unheard but imagined five times as well as others. Several distinctions between the paradigms help between robust false misinformation memories and false to explain limits of inter-paradigm relationships. DRM Performed judgments to items listened-to in Session 1 false memories are the result of spontaneous internal pro- and later imagined once or five times. False memories in cesses, misinformation memories arise through external both paradigms may arise from source monitoring failures suggestion, and imagination inflation develops when through external suggestion (misinformation) or direction what was imagined confounds what was real. Imagin- (imagination). Though speculative, this could account for ation-inflated memories are autobiographical in nature, the positive correlations revealed here. However, the whereas misinformation memories concern an external number and magnitude of these three correlations event, and the DRM produces false memories more so should not be overstated; the 21 other possible corre- through semantic rather than episodic memory stores. lations between false memories are not significant. The DRM also spans two crucial phases (encoding and Other correlations emerged between DRM false mem- test) while the misinformation and imagination paradigms ories and imagined false memories. Specifically, total span three (encoding, suggestion/imagination, and test). false recognition of lures was associated with Performed These tasks are quite different in terms of their methods judgments to items listened to in Session 1 and later ima- but also in how they are ultimately experienced by the par- gined. Robustly remembered lures, on the other hand, ticipant. When considering whether some people may were associated with false Heard judgments and false Per- have a “false memory trait”, it is important also to consider formed judgments across every major category of item (see that the dissimilarities between these paradigms may Table 4). Because robustly recognised lures are those for simply trump the fact that they all produce false memories which participants report an actual memory of having of one kind or another (consistent with Bernstein et al.’s seen the item, rather than a mere belief that they saw it, [2018] article title, “‘False Memory’ is a linguistic these particular memories are better explained by acti- convenience”). vation-monitoring theories than are total false recognition of the lures. Source monitoring failures also likely account Individual differences for false memories arising from imagination inflation; par- ticipants may develop these memories when they cannot A large number of individual differences were examined in accurately identify the source of their memory as their this study as potential predictors of false memories. As the Session 2 imagination instead of that which occurred in number of comparisons increases, so too does the prob- Session 1. From a monitoring perspective, therefore, it is ability of committing Type I errors (false positive corre- not surprising to find more relationships between ima- lations). Nevertheless, relatively few of these measures gined memories and robust false recognition of lures predict memory – veridical or illusory – in the three para- than imagined memories and total false recognition of digms. When taken together in an ordinary least squares lures. However, a test of shared monitoring failures multiple regression analysis, the 19 predictors do not amongst the paradigms was not a component of this account for a significant amount of the variance in the pro- study. portion of misinformation items recognised (p = .61) critical Remarkably, confidence judgments for false memory lures recognised (p = .09), or false Heard judgments to items in all three paradigms were correlated. These corre- unheard, imagined items (p = .33), despite large sample lations are small between judgments for misinformation sizes for each paradigm (n = 244–297 after attrition) and items and both critical lures and imagination items, but no problems with collinearity or residual diagnostics. the correlation between confidence judgments for critical Therefore, each of the predictors is discussed in the lures and imagined items was strikingly large (r = .59). context of individual paradigms. The correlation between confidence for studied items in In the misinformation task, there are zero significant the DRM and heard but unimagined items in the imagin- individual difference correlates of false memory. These ation inflation paradigm was also correlated, though to a include failures to replicate relationships founds in the lit- far lesser degree. The most obvious similarities between erature between misinformation false memories and vivid- DRM judgments and imagination inflation judgments are ness of visual imagery, the agreeableness and openness that they involve a yes-no recognition test, while the mis- subscales of the Big Five personality inventory, and disso- information paradigm does not, and both tests occurred ciative experiences. This may be due in part to methodo- during Session 3 while the misinformation paradigm took logical and analytical differences between the reported 978 R. M. NICHOLS AND E. F. LOFTUS studies and this one. For example, Tomes and Katz (1997) with a decrease in OCQ bias scores and an increase in dis- compared Vividness of Visual Imagery Questionnaire sociativity. This latter relationship confirms the results of scores between participants who endorsed all three misin- previous studies (Heaps & Nash, 1999; Hyman & Billings, formation items in their design to those who endorsed 1998). False Performed judgments here are also associated fewer than three misinformation items. Though dichotomi- with absorption and anomalous experiences. Absorption sation of continuous variables may reveal group differ- has been shown to be associated with an increase in ences where none exist in correlational analyses, there more naturally occurring false autobiographical memories are several limits of this approach that result in more (Platt et al., 1998), and those who are more prone to liberal analysis (Irwin & McClelland, 2003; Maxwell & absorption in their environments may be more suggestible Delaney, 1993). Furthermore, Eisen and Carlson (1998) in a variety of contexts, imagination inflation notwithstand- report significant positive correlations between misinfor- ing. Those who report having more anomalous experiences mation errors and both dissociative experiences and may also be prone to developing false autobiographical absorption, but these correlations are significant at the p memories because the anomalous experiences they < .05 level only with one-tailed tests. All analyses con- report may themselves be confabulations. Interestingly, ducted in this investigation were calculated with two- anomalous experiences are associated only with false Per- tailed tests given adequate statistical power to detect formed judgments, not with false Heard judgments; differences (i.e., post-hoc power analyses revealed Power perhaps reports of events that are as extreme and unlikely [1 – β error] = .99 to detect a medium-sized correlation as those listed in the AEI (e.g., “At times, I have felt pos- between absorption and misinformation errors in this sessed by an outside force”) are associated only with study). reports as extreme and unlikely as performing an action In the DRM, the Cognitive Reflection Task is most highly statement that was not even encountered in Session 1. correlated with the proportion of lures falsely remembered. To summarise, several individual difference variables More specifically, the total sum of intuitive answers is posi- were investigated as predictors of false memory in the mis- tively correlated with false memories while the total sum of information, DRM, and imagination inflation paradigms. correct answers is negatively correlated. These relation- Very few are significantly correlated with false memories, ships may be explained by the fact that the CRT is designed and those that did correlate represent small and small-to- to lure respondents into providing an intuitive but incor- medium sized effects. The only common predictor to rect answer, much in the way that the DRM is designed false memories in multiple paradigms is one’s report of to lure respondents into endorsing intuitive but never pre- anomalous experiences, and this predicts both robust sented critical items at test. Conversely, the ability to sup- false memories in the DRM and false Performed judgments press an intuitive, System 1 (Stanovich & West, 2000) to unheard, imagined items in the imagination inflation thought process to arrive at correct answer in the CRT paradigm to the same degree (r = .13). The relative inability may be related to the ability to resist endorsement of a of these variables to predict false memories is consistent critical lure. Frederick (2005) also describes his CRT as a with the Dual Encoding Interference hypothesis (Patihis, test that requires cognitive ability in order to generate 2018). This hypothesis posts that in a misinformation correct answers; the negative correlation between correct task, traits that enhance encoding will support encoding answers and critical lures is therefore compatible with of both true and false memories. Similarly, traits that findings that suggest cognitive ability, measured through weaken encoding will result in both weak misinformation a variety of tasks, is negatively correlated with false mem- memory and weak memory for original event details. ories in the DRM (Zhu et al., 2010b). Surprisingly, OCQ dis- Both types of traits will produce a moderate probability crimination (a proxy for cognitive ability in this study) is not of false memory that might be akin to someone that significantly associated with false memories. Research has possess neither high nor very low amounts of the trait, shown that working memory capacity is somewhat protec- resulting in a null correlation. This may be particularly tive against the development of DRM errors (Watson et al., applicable to other false memories involving source moni- 2005), but discrimination scores on the OCQ discrimination toring failures like those arising from imagination inflation. depend on a subject’s familiarity with pieces of general, The relevance of this hypothesis to DRM false memories is trivial-like knowledge rather than working memory limited, however, as DRM lures are not encoded at the time capabilities. of study. Finally, one’s propensity to develop false memories after imagination is correlated with a few of the individual differ- Practical implications ence variables. For example, both an increase in false Heard and false Performed judgments are associated with a If it is the case that we are all vulnerable to memory distor- decrease in agreeableness. Though some have decried tion, then this has important implications for both clinical misinformation effects as an acquiescence to demand and legal practitioners. In a court case, for example, an characteristics or others social pressures (McCloskey & Zar- expert witness (whom we will call Dr. B) denounced the agoza, 1985), this finding does not support that claim. False idea that an alleged abuse victim was likely to have experi- Heard and false Performed judgments are also associated enced false memories about her abuse given her low MEMORY 979 scores on the Gudjonsson Suggestibility Scale (Doe v. Hart- DRM (distractor items) and the imagination inflation para- ford Roman Catholic, 2014). This scale (Gudjonsson, 1984) digm (unheard, unimagined items). This study lacks the employs a misinformation-type methodology, with an orig- ability to compare endorsement rates of novel items inal event described in a narrative, to gauge how willing among all three paradigms. In a similar vein, the imagin- people are both to yield to suggestive questions and to ation inflation test did not contain a measure of robust shift their initial responses in the face of social pressure. false memory such as the source judgments for misinfor- To date, there is no published evidence to indicate that mation items or, more relevant to imagination inflation, scores on this scale predict other types of false memories, remember/know judgments for DRM items. This limits to just as misinformation false memories are not strongly cor- ability to interpret any false heard, listened, or performed related with DRM or imagination false memories in this judgment as arising from false memory rather than false study. A clinician who wishes to follow Dr. B’s approach, belief and prevents comparing memories with more recol- however, may feel liberty to employ a controversial thera- lective qualities across all three tasks. peutic technique because his or her client has a low GSS Another limitation of this study is that participants did score, recognises relatively few critical DRM lures, does not have the opportunity to provide open-ended not seem to incorporate imaginations into her memory responses. This means that information about the partici- for actual events, or through other tests, does not seem pants’ recollective experiences is limited to remember/ to produce false memories to a large degree. This hypothe- know judgments for DRM items and performed/ima- tical situation is reckless if not completely dangerous. Trial gined/listened-to judgments for imagined items that outcomes may be unfairly influenced and people’s mem- were judged as Heard. Information about a rememberer’s ories may be outright damaged by practices that remain subjective experiences is central component of false wholly unsupported by science. memory theories, and some have claimed that false mem- ories cannot be entirely understood without considering them (Lampinen, Neuschatz, & Payne, 1997). Future Strengths, limitations and future directions research would benefit from allowing participants to This study benefits from a very large sample size for a freely describe their subjective phenomenology in multiple study whose procedures occurred in the laboratory span- paradigms; content analysis could then shed light on ning three separate sessions over two entire weeks. Fortu- how experience allows researchers to describe all three nately, the three paradigms utilised here produced false memory phenomena as false memory but also help elabor- memories as expected. This, in conjunction with the ate upon the small to null relationships discovered in this large sample size, minimises the likelihood that null corre- research. lations are the result of Type II (false negative) errors. This Another issue concerns the internal reliability of compo- within-subjects false memory study also includes imagin- site measures used in this study. It is possible that relation- ation inflation as a measure of verifiable, false autobiogra- ships between variables are not detected to the extent that phical memory. It is the first to suggest minimal to that the variables themselves do not have high internal nonexistent relationships between imagined false mem- reliability (Spearman, 1904). While internal consistency ories and misinformation and DRM false memories. was sufficient for the individual difference measures (see Unlike uncorroborated false memories arising from Table 5) and DRM memories (α = .67 for critical items, α some types of false event suggestion, the false memories = .82 for studied items α = .91 for distractor items), in this study are fully known to be false and therefore offer reliability estimates for false memories in the misinforma- a methodological advantage. Finally, whether a false tion and imagination inflation tasks could not be calculated memory trait exists should be addressed not just by due to item counterbalancing. However, the reliabilities examining relationships between false memories them- reported here are comparable or higher to those reported selves, but also by identifying other factors that represent by Ost et al. (2013), who noted that attenuation due to shared variance among them. This multiple paradigms limited reliability would have rendered the correlations study also includes the largest battery of individual differ- smaller but still discoverable. ence predictors to date. In addition to addressing these limitations, future However, limitations of this study must be addressed. research may benefit from addressing the question of First, memory assessment varies among the three para- trait susceptibility through other psychological lenses. digms. For example, the misinformation test contained One of these is the use of neuroimaging. Functional MRI two-alternative, forced-choice recognition questions (fMRI) studies have demonstrated meaningful patterns of while the DRM and imagination inflation test items brain activity for true and false memories in both the mis- required yes-no recognition judgments. The misinforma- information paradigm (Baym & Gonsalves, 2010; Okado & tion tests lacked a third alternative: a novel item that was Stark, 2005; Stark, Okado, & Loftus, 2010) and the DRM not presented in either the original event nor in the misin- (Cabeza, Rao, Wagner, Mayer, & Schacter, 2001; Schacter, formation narrative. The absence of this item may have Buckner, Koutstaal, Dale, & Rosen, 1997), yet no one slightly inflated the misinformation effect, but more impor- study to date has compared structural or functional brain tantly, similar control items were present in the both the differences in participants who complete multiple false 980 R. M. NICHOLS AND E. F. LOFTUS memory tasks in the same study. This biological perspec- resistant, to incorporating all different kinds of those tive may shed more light on the patterns of data discov- extra bits and pieces into memory. ered here. Furthermore, given the few, small correlations uncov- ered between personality variables and false memories Data availability statement in this study, researchers who are interested in further The data that support the findings of this study are avail- exploring the relationships between false memory para- able on request from the corresponding author, RMN. digms may be interested in taking an experimental The data are not publicly available due to their containing ff rather than an individual di erences approach. For information that could compromise the privacy of research ff instance, researchers have studied the e ects of divided participants. attention on the production of false memories. In the DRM, dividing participants’ attention at encoding has been show to decrease false recognition of critical lures Disclosure statement while dividing their attention at retrieval has been No potential conflict of interest was reported by the authors. shown to increase critical lures (Knott & Dewhurst, 2007). In the misinformation task, participants whose attention is divided at both encoding (Lane, 2006)and Funding retrieval (Zaragoza & Lane, 1998) are more likely to incor- porate misleading suggestions into memory. Little is This work was supported by an American Psychology-Law Society (APLS; APA Division 41) Grant-in-Aid. known about how divided attention affects the develop- ment of false autobiographical memories, and nothing ff ff is known about how it a ects di erent types of false References memories within subjects. To the degree that divided attention or any other manipulation affects false mem- Balota, D. A., Cortese, M. J., Duchek, J. M., Adams, D., Roediger, H. L., McDermott, K. B., & Yerys, B. E. (1999). Veridical and false memory ories in the same manner, more can be said about the in healthy older adults and in Dementia of the Alzheimer’s type. mechanisms shared by those paradigms. Cognitive Neuropsychology, 16, 361–384. Baym, C. L., & Gonsalves, B. D. (2010). Comparison of neural activity that leads to true memories, false memories, and forgetting: An fMRI study of the misinformation effect. Cognitive, Affective, & Conclusion Behavioral Neuroscience, 10(3), 339–348. doi:10.3758/CABN.10.3.339 Belli, R. F., & Loftus, E. F. (1994). Recovered memories of childhood The research question explored here is simply defined, fun- abuse: A source monitoring perspective. In S. J. Lynn, & J. Rhue damental to understanding memory, and contributes to a (Eds.), Dissociation: Theory, clinical, and research perspectives (pp. small but growing literature on an understudied question: 415–433). New York: Guilford. are there certain people who are prone to developing false Bernstein, D. M., Laney, C., Morris, E. K., & Loftus, E. F. (2005a). False beliefs about fattening foods can have healthy consequences. memories across multiple contexts? In summary, it seems Proceedings of the National Academy of Sciences, 102, 13724–13731. that the answer is no. The results of this one large study Bernstein, D. M., Laney, C., Morris, E. K., & Loftus, E. F. (2005b). False fail to discover any substantial relationships between false memories about food can lead to food avoidance. Social memories in three classic experimental paradigms: the mis- Cognition, 23,11–34. “ information paradigm, the DRM, and imagination inflation. Bernstein, D. M., Scoboria, A., Desjarlais, L., & Soucie, K. (2018). False memory” is a linguistic convenience. Psychology of Consciousness: There were small to null relationships between indices of Theory, Research, and Practice, 5(2), 161–179. doi:10.1037/ false memory among the three tasks, and while some indi- cns0000148 vidual differences predict false memories to a small Blair, I., Lenton, A., & Hastie, R. (2002). The reliability of the DRM para- degree, not one of the personality variables explored in digm as a measure of individual differences in false memories. – this study predicts false memories in all three contexts. Psychonomic Bulletin & Review, 9(3), 590 596. doi:10.3758/ BF03196317 These data depict the ubiquitous nature of memory dis- Bottoms, B. L., & Davis, S. L. (1997). The creation of satanic ritual abuse. tortion. And even though the memories discussed here Journal of Social and Clinical Psychology, 16, 112–132. were produced in the lab, real-world memories may be Brainerd, C. J., & Reyna, V. F. ( 2005). The science of false memory. New influenced through the same means. When we attempt York, NY: Oxford University Press. to remember and therefore reconstruct our memories, Bremner, J. D., Shobe, K. K., & Kihlstrom, J. F. (2000). False memories in women with self-reported childhood sexual abuse: An empirical those memories may contain extra bits and pieces from study. Psychological Science, 11, 333–337. what we have heard, read, or seen from other sources; Brewin, C. R., & Andrews, B. (2014). Why it is scientifically respectable to from what seems to “fit” based on schematic congruence; believe in repression: A response to Patihis, Ho, Tingen, Lilienfeld, or even from details generated through our own internal and Loftus (2014). Psychological Science, 25(10), 1964–1966. replays and imaginations. Importantly, this research doi:10.1177/0956797614541856 Butler, K. M., McDaniel, M. A., Dornburg, C. C., Price, A. L., & Roediger, H. demonstrates that distortions can worm their way into L. (2004). Age differences in veridical and false recall are not inevi- very different kinds of memories for us all, and no one table: The role of frontal lobe function. Psychonomic Bulletin & type of person is particularly vulnerable, nor particularly Review, 11, 921–925. MEMORY 981

Cabeza, R., & Lennartson, R. (2005). False memory across languages: Gallo, D. A. (2006). Associative illusions of memory: False memory Implicit associative response vs fuzzy trace views. Memory, 13(1), research in DRM and related tasks. New York: Psychology Press. 1–5. doi:10.1080/09658210344000161 Gallo, D. A. (2010). False memories and fantastic beliefs: 15 years of the Cabeza, R., Rao, S. M., Wagner, A. D., Mayer, A. R., & Schacter, D. L. DRM illusion. Memory & Cognition, 38(7), 833–848. doi:10.3758/MC. (2001). Can medial temporal lobe regions distinguish true from 38.7.833 false? An event-related functional MRI study of veridical and illusory Gallo, D. A., Bell, D. M., Beier, J. S., & Schacter, D. L. (2006). Two types recognition memory. Proceedings of the National Academy of of recollection-based monitoring in younger and older adults: Sciences, 98(8), 4805–4810. doi:10.1073/pnas.081082698 Recall-to-reject and the distinctiveness heuristic. Memory, 14, Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of 730–741. Personality and Social Psychology, 42(1), 116–131. Garry, M., Manning, C. G., Loftus, E. F., & Sherman, S. J. (1996). Calvillo, D. P. (2014). Individual differences in susceptibility to misinfor- Imagination inflation: Imagining a childhood event inflates confi- mation effects and hindsight bias. The Journal of General dence that it occurred. Psychonomic Bulletin & Review , 3(2), 208–214. Psychology, 141(4), 393–407. doi:10.1080/00221309.2014.954917 Geraerts, E., Lindsay, D. S., Merckelbach, H., Jelicic, M., Raymaekers, L., Calvillo, D. P., & Parong, J. A. (2016). The misinformation effect is unre- Arnold, M. M., & Schooler, J. W. (2009). Cognitive mechanisms lated to the DRM effect with and without a DRM warning. Memory, underlying recovered-memory experiences of childhood sexual 24(3), 324–333. doi:10.1080/09658211.2015.1005633 abuse. Psychological Science, 20(1), 92–98. doi:10.1111/j.1467-9280. Calvillo, D. P., Vasquez, A. N., & Pesavento, A. (2018). Imagination 2008.02247.x inflation effects are unrelated across two imagination inflation Geraerts, E., Smeets, E., Jelicic, M., van Heerden, J., & Merckelbach, H. tasks. Psychology of Consciousness: Theory, Research, and Practice. (2005). Fantasy proneness, but not self-reported trauma is related Advance online publication. doi:10.1037/cns0000178 to DRM performance of women reporting recovered memories of Cann, D. R., & Katz, A. N. (2005). Habitual acceptance of misinformation: childhood sexual abuse. Consciousness and Cognition, 14(3), 602– Examination of individual differences and source attributions. 612. doi:10.1016/j.concog.2005.01.006 Memory and Cognition, 33(3), 405–417. Goff, L. M., & Roediger III, H. L. (1998). Imagination inflation for action Clancy, S. A., McNally, R. J., Schacter, D. L., Lenzenweger, M. F., & events: Repeated imaginings lead to illusory recollections. Memory& Pitman, R. K. (2002). Memory distortion in people reporting abduc- Cognition, 26,20–33. tion by aliens. Journal of Abnormal Psychology, 111 , 455–461. Gorsuch, R. L., & Venable, G. D. (1983). Development of an “age univer- Clancy, S. A., Schacter, D. L., McNally, R. J., & Pitman, R. K. (2000). False sal” I-E scale. Journal for the Scientific Study of Religion, 22(2), 181– recognition in women reporting recovered memories of sexual 187. doi:10.2307/1385677 abuse. Psychological Science, 11,26–31. Graham, L. M. (2007). Need for cognition and false memory in the Corson, Y., Verrier, N., & Bucic, A. (2009). False memories and individual Deese-Roediger-McDermott paradigm. Personality and Individual variations: The role of field dependence-independence. Personality Differences, 42(3), 409–418. doi:10.1016/j.paid.2006.07.012 and Individual Differences, 47,8–11. Gudjonsson, G. H. (1984). A new scale of interrogative suggestibility. Cui, X., Jeter, C. B., Yang, D., Montague, P. R., & Eagleman, D. M. (2007). Personality and Individual Differences, 5(3), 303–314. doi:10.1016/ Vividness of mental imagery: Individual variability can be measured 0191-8869(84)90069-2 objectively. Vision Research, 47, 474–478. doi:10.1016/j.visres.2006. Heaps, C., & Nash, M. (1999). Individual differences in imagination 11.013 inflation. Psychonomic Bulletin & Review, 6, 313–318. doi:10.3758/ Deese, J. (1959). On the prediction of occurrence of particular verbal BF03214120 intrusions in immediate recall. Journal of Experimental Psychology, Hekkanen, S. T., & McEvoy, C. (2002). False memories and source-moni- 58,17–22. toring problems: Criterion differences. Applied Cognitive Psychology, Dehon, H., Bastin, C., & Larøi, F. (2008). The influence of delusional idea- 16(1), 73–85. doi:10.1002/acp.753 tion and dissociative experiences on the resistance to false mem- Horselenberg, R., Merckelbach, H., Muris, P., Rassin, E., Sijsenaar, M., & ories in normal healthy subjects. Personality and Individual Spaan, V. (2000). Imagining fictitious childhood events: The role Differences, 45(1), 62–67. doi:10.1016/j.paid.2008.02.020 of individual differences in imagination inflation. Clinical Doe v. Hartford Roman Catholic et al., No. X10-UWY-CV10-5015963-S. Psychology & Psychotherapy, 7(2), 128–137. (2014). Transcript of deposition of Daniel P. Brown, Ph.D., March Horselenberg, R., Merckelbach, H., van Breukelen, G., & Wessel, I. (2004). 10, 2014. Superior Court of Waterbury, CT. Individual differences in the accuracy of autobiographical memory. Eisen, M. L., & Carlson, E. B. (1998). Individual differences in suggestibil- Clinical Psychology & Psychotherapy, 11(3), 168–176. doi:10.1002/ ity: Examining the influence of dissociation, absorption, and a cpp.400 history of childhood abuse. Applied Cognitive Psychology, 12(7), Hyman, J., & Billings, F. J. (1998). Individual differences and the creation S47–S61. doi:10.1002/(SICI)1099-0720(199812)12:7<S47::AID- of false childhood memories. Memory, 6(1), 1–20. doi:10.1080/ ACP598>3.0.CO;2-P 096582198388373 Eisen, M. L., Gomes, D. M., Lorber, W. G., Perez, C. I., & Uchishiba, H. Irwin, J. R., & McClelland, G. H. (2003). Negative consequences of (2013). Using an individual differences approach to examine two dichotomizing continuous predictor variables. Journal of distinct types of suggestibility effects. Applied Cognitive Marketing Research, 40(3), 366–371. doi:10.1509/jmkr.40.3.366. Psychology, 27(1), 2–11. doi:10.1002/acp.2864 19237 Eisen, M. L., Morgan, D. Y., & Mickes, L. (2002). Individual differences in Jacoby, L., Kelley, C., & Dywan, J. (1989). Memory attributions. In eyewitness memory and suggestibility: Examining relations Varieties of memory and consciousness: Essays in honour of Endel between acquiescence, dissociation and resistance to misleading Tulving (pp. 391–422). Hillsdale, NJ: Erlbaum. information. Personality and Individual Differences, 33(4), 553–571. Jaschinski, U., & Wentura, D. (2002). Misleading postevent information doi:10.1016/S0191-8869(01)00172-6 and working memory capacity: An individual differences approach Epstein, S., Pacini, R., DenesRaj, V., & Heier, H. (1996). Individual differ- to eyewitness memory. Applied Cognitive Psychology, 16(2), 223– ences in intuitive-experiential and analytical-rational thinking 231. doi:10.1002/acp.783 styles. Journal of Personality and Social Psychology, 71(2), 390–405. John, O. P., Donahue, E. M., & Kentle, R. L. (1991 ). The Big Five Inventory-- Frederick, S. (2005). Cognitive reflection and decision making. Journal Versions 4a and 54. Berkeley, CA: University of California, Berkeley, of Economic Perspectives, 19(4), 25–42. Institute of Personality and Social Research. Gallagher, C., Kumar, V. K., & Pekala, R. J. (1994). The anomalous experi- John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to the ences inventory: Reliability and validity. Journal of Parapsychology, integrative Big Five trait taxonomy: History, measurement, and con- 58(4), 402–428. ceptual issues. In O. P. John, R. W. Robins, & L. A. Pervin (Eds.), 982 R. M. NICHOLS AND E. F. LOFTUS

Handbook of personality: Theory and research (pp. 114–158). Meyersburg, C. A., Bogdan, R., Gallo, D. A., & McNally, R. J. (2009). False New York, NY: Guilford Press. memory propensity in people reporting recovered memories of Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source past lives. Journal of Abnormal Psychology, 118(2), 399–404. monitoring. Psychological Bulletin, 114(1), 3–28. doi:10.1037/0033- doi:10.1037/a0015371 2909.114.1.3 Mitchell, K. J., & Johnson, M. K. (2000). Source monitoring: Attributing Kensinger, E. A., & Schacter, D. L. (1999). When true memories suppress mental experiences. In E. Tulving & F. I. M. Craik (Eds.), The Oxford false memories: Effects of aging. Cognitive Neuropsychology, 16, handbook of memory (pp. 179–195). New York, NY: Oxford 399–415. University Press. Knott, L. M., & Dewhurst, S. A. (2007). The effects of divided attention at Monds, L. A., Paterson, H. M., & Kemp, R. I. (2017). Do emotional stimuli study and test on false recognition: A comparison of DRM and cate- enhance or impede recall relative to neutral stimuli? An investi- gorized lists. Memory & Cognition, 35(8), 1954–1965. doi:10.3758/ gation of two “false memory” tasks. Memory, 25(8), 945–952. BF03192928 doi:10.1080/09658211.2016.1237653 Lampinen, J. M., Neuschatz, J. S., & Payne, D. G. (1997). Memory illu- Nash, R. A., & Wade, K. A. (2009). Innocent but proven guilty: Eliciting sions and consciousness: Examining the phenomenology of true internalized false confessions using doctored-video evidence. and false memories. Current Psychology, 16(3-4), 181–224. Applied Cognitive Psychology, 23, 624–637. doi:10.1002/acp.1500 Lane, S. M. (2006). Dividing attention during a witnessed event Nash, R. A., Wade, K. A., & Lindsay, D. S. (2009). Digitally manipulating increases eyewitness suggestibility. Applied Cognitive Psychology, memory: Effects of doctored videos and imagination in distorting 20(2), 199–212. doi:10.1002/acp.1177 beliefs and memories. Memory & Cognition, 37(4), 414–424. doi:10. Laney, C., & Loftus, E. F. (2008). Emotional content of true and false 3758/MC.37.4.414 memories. Memory, 16, 500–516. doi:10.1080/09658210802065939 Nichols, R. M., & Loftus, E. F. (2009). Do cognitive styles predict false Laney, C., Morris, E. K., Bernstein, D. M., Wakefield, B. M., & Loftus, E. F. memory susceptibility? A two-paradigm analysis. Unpublished (2008). Asparagus, a love story: Healthier eating could be just a false raw data. belief away. Experimental Psychology, 55, 291–300. Okado, Y., & Stark, C. E. L. (2005). Neural activity during encoding pre- Laws, K. R., & Bhatt, R. (2005). False memories and delusional ideation dicts false memories created by misinformation. Learning & in normal healthy subjects. Personality and Individual Differences, 39 Memory, 12(1), 3–11. doi:10.1101/lm.87605 (4), 775–781. doi:10.1016/j.paid.2005.03.005 Oppenheimer, D. M., Meyvis, T., & Davidenko, N. (2009). Instructional Leding, J. K. (2011). Need for Cognition and false recall. Personality and manipulation checks: Detecting satisficing to increase statistical Individual Differences, 51(1), 68–72. doi:10.1016/j.paid.2011.03.017 power. Journal of Experimental Social Psychology, 45(4), 867–872. Liebman, J. I., McKinley-Pace, M. J., Leonard, A. M., Sheesley, L. A., doi:10.1016/j.jesp.2009.03.009 Gallant, C. L., Renkey, M. E., & Lehman, E. B. (2002). Cognitive and Ost, J., Blank, H., Davies, J., Jones, G., Lambert, K., Salmon, K., & Lu, L. psychosocial correlates of adults’ eyewitness accuracy and suggest- (2013). False memory ≠ false memory: DRM errors are unrelated ibility. Personality and Individual Differences, 33(1), 49–66. doi:10. to the misinformation effect. PLoS ONE, 8(4), e57939. doi:10.1371/ 1016/S0191-8869(01)00135-0 journal.pone.0057939 Loftus, E. F. (2017). Eavesdropping on memory. Annual Review of Otgaar, H., & Candel, I. (2011). Children’s false memories: Different false Psychology, 68(1), 1–18. doi:10.1146/annurev-psych-010416-044138 memory paradigms reveal different results. Psychology, Crime & Loftus, E. F., & Bernstein, D. M. (2005). Rich false memories. In A. F. Healy Law, 17(6), 513–528. doi:10.1080/10683160903373392 (Ed.), Experimental cognitive psychology and its applications (pp. Otgaar, H., Verschuere, B., Meijer, E. H., & van Oorsouw, K. (2012). The 101–113). Washington, DC: American Psychological Association. origin of children’ s implanted false memories: Memory traces or Loftus, E. F., & Mazzoni, G. A. L. (1998). Using imagination and person- compliance? Acta Psychologica, 139(3), 397–403. doi:10.1016/j.actpsy alized suggestion to change people. Behavior Therapy, 29, 691–706. Paddock, J. R., Terranova, S., Kwok, R., & Halpern, D. V. (2000). When Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of knowing becomes remembering: Individual differences in suscepti- verbal information into a visual memory. Journal of Experimental bility to suggestion. The Journal of Genetic Psychology, 161(4), 453– Psychology: Human Learning and Memory, 4,19–31. 468. Loftus, E. F., & Pickrell, J. E. (1995). The formation of false memories. Parker, S., Garry, M., Engle, R. W., Harper, D. N., & Clifasefi,S.L.(2008). Psychiatric Annals, 25, 720–725. Psychotropic placebos reduce the misinformation effect by increas- Marks, D. F. (1973). Visual imagery differences in the recall of pictures. ing monitoring at test. Memory, 16(4), 410–419. doi:10.1080/ British Journal of Psychology, 64(1), 17–24. doi:10.1111/j.2044-8295. 09658210801956922 1973.tb01322.x Patihis, L. (2018). Why there is no false memory trait and why everyone Matthews, G. A., & Harley, T. A. (1993). Effects of extraversion and self- is susceptible to memory distortions: The dual encoding interfer- report arousal on semantic priming: A connectionist approach. ence hypothesis (Commentary on Bernstein, Scoboria, Desjarlais, Journal of Personality and Social Psychology, 65(4), 735–756. & Soucie, 2018). Psychology of Consciousness: Theory, Research, doi:10.1037/0022-3514.65.4.735 and Practice, 5(2), 180–184. doi:10.1037/cns0000143 Maxwell, S. E., & Delaney, H. D. (1993). Bivariate median splits and spur- Patihis, L., Frenda, S. J., & Loftus, E. F. (2018). False memory tasks do not ious statistical significance. Psychological Bulletin, 113(1), 181–190. reliably predict other false memories. Psychology of Consciousness: doi:10.1037/0033-2909.113.1.181 Theory, Research, and Practice, 5(2), 140–160. Mazzoni, G. A. L., Lombardo, P., Malvagia, S., & Loftus, E. F. (1999). Patihis, L., Ho, L. Y., Tingen, I. W., Lilienfeld, S. O., & Loftus, E. F. (2014). Dream interpretation and false beliefs. Professional Psychology: Are the “memory wars” over? A scientist-practitioner gap in beliefs Research and Practice, 30(1), 45–50. doi:10.1037/0735-7028.30.1.45 about repressed memory. Psychological Science, 25(2), 519–530. Mazzoni, G., & Memon, A. (2003). Imagination can create false autobio- doi:10.1177/0956797613510718 graphical memories. Psychological Science, 14(2), 186–188. Patihis, L., Lilienfeld, S. O., Ho, L. Y., & Loftus, E. F. (2014). Unconscious McCloskey, M., & Zaragoza, M. (1985). Misleading postevent infor- repressed memory is scientifically questionable. Psychological mation and memory for events: Arguments and evidence against Science, 25(10), 1967–1968. doi:10.1177/0956797614547365 memory impairment hypotheses. Journal of Experimental Paulhus, D. L., & Harms, P. D. (2004). Measuring cognitive ability with Psychology: General, 114(1), 1–16. doi:10.1037/0096-3445.114.1.1 the over-claiming technique. Intelligence, 32, 297–314. Merckelbach, H., Horselenberg, R., & Muris, P. (2001). The Creative Paulhus, D. L., Harms, P. D., Bruce, M. N., & Lysy, D. C. (2003). The Experiences Questionnaire (CEQ): A brief self-report measure of over-claiming technique: Measuring self-enhancement indepen- fantasy proneness. Personality and Individual Differences, 31(6), dent of ability. Journal of Personality and Social Psychology, 84, 987–995. doi:10.1016/S0191-8869(00)00201-4 890–904. MEMORY 983

Peters, M. J. V., Jelicic, M., Haas, N., & Merckelbach, H. (2006). Mild Spearman, C. (1904). The proof and measurement of association executive dysfunctions in undergraduates are related to recollect- between two things. The American Journal of Psychology, 15(1), ing words never presented. International Journal of Neuroscience, 72–101. doi:10.2307/1412159 116(9), 1065–1077. doi:10.1080/00207450600808768 Stadler, M. A., Roediger, H. L., & McDermott, K. B. (1999). Norms for Peters, M. J. V., Jelicic, M., Verbeek, H., & Merckelbach, H. (2007). Poor word lists that create false memories. Memory and Cognition, 27, working memory predicts false memories. European Journal of 494–500. Cognitive Psychology, 19(2), 213–232. doi:10.1080/09541440600 Stanovich, K. E., & West, R. F. (2000). Individual differences in reasoning: 760396 Implications for the rationality debate? Behavioral and Brain Peters, M. J. V., Smeets, T., Giesbrecht, T., Jelicic, M., & Merckelbach, H. Sciences, 23(5), 645–665. (2007). Confusing action and imagination. The Journal of Nervous Stark, C. E. L., Okado, Y., & Loftus, E. F. (2010). Imaging the reconstruc- and Mental Disease, 195(9), 752–757. doi:10.1097/NMD. tion of true and false memories using sensory reactivation and the 0b013e318142cc02 misinformation paradigms. Learning & Memory, 17(10), 485–488. Pezdek, K., & Salim, R. (2011). Physiological, psychological and behav- doi:10.1101/lm.1845710 ioral consequences of activating autobiographical memories. Sternberg, R. J. (1997). Thinking styles. New York: Cambridge University Journal of Experimental Social Psychology, 47(6), 1214–1218. Press. doi:10.1016/j.jesp.2011.05.004 Strange, D., Sutherland, R., & Garry, M. (2006). Event plausibility does Platt, R. D., Lacey, S. C., Iobst, A. D., & Finkelman, D. (1998). Absorption, not determine children’s false memories. Memory, 14(8), 937–951. dissociation, and fantasy-proneness as predictors of memory distor- doi:10.1080/09658210600896105 tion in autobiographical and laboratory-generated memories. Takarangi, M. K. T., Parker, S., & Garry, M. (2006). Modernising the mis- Applied Cognitive Psychology, 12(7), S77–S89. doi:10.1002/ information effect: the development of a new stimulus set. Applied (SICI)1099-0720(199812)12:7<S77::AID-ACP601>3.0.CO;2-C Cognitive Psychology, 20(5), 583–590. doi:10.1002/acp.1209 Porter, S., Birt, A. R., Yuille, J. C., & Lehman, D. R. (2000). Negotiating Tellegen, A., & Atkinson, G. (1974). Openness to absorbing and self- false memories: Interviewer and rememberer characteristics relate altering experiences (“absorption”), a trait related to hypnotic sus- to memory distortion. Psychological Science, 11(6), 507–510. ceptibility. Journal of Abnormal Psychology , 83(3), 268–277. doi:10. doi:10.1111/1467-9280.00297 1037/h0036681 Putnam, F. W. (1995). Development of dissociative disorders. In D. Thomas, A. K., Bulevich, J. B., & Loftus, E. F. (2003). Exploring the role of Cicchetti, & D. J. Cohen (Eds.), Developmental psychology, volume 2: repetition and sensory elaboration in the imagination inflation Risks, Disorders, and Adaption (pp. 581–608). New York: Wiley. effect. Memory & Cognition, 31(4), 630–640. Raymaekers, L., Peters, M. J. V., Smeets, T., Abidi, L., & Merckelbach, H. Thomas, A. K., & Loftus, E. F. (2002). Creating bizarre false memories (2011). Underestimation of prior remembering and susceptibility to through imagination. Memory & Cognition, 30(3), 423–431. doi:10. false memories: Two sides of the same coin? Consciousness and 3758/BF03194942 Cognition, 20(4), 1144–1153. doi:10.1016/j.concog.2010.12.010 Tomes, J. L., & Katz, A. N. (1997). Habitual susceptibility to misinforma- Reyna, V. F., Corbin, J. C., Weldon, R. B., & Brainerd, C. J. (2016). How tion and individual differences in eyewitness memory. Applied fuzzy-trace theory predicts true and false memories for words, sen- Cognitive Psychology, 11(3), 233–251. doi:10.1002/(SICI)1099-0720 tences, and narratives. Journal of Applied Research in Memory and (199706)11:3<233::AID-ACP447>3.0.CO;2-V Cognition, 5(1), 1–9. doi:10.1016/j.jarmac.2015.12.003 Tomes, J. L., & Katz, A. N. (2000). Confidence-accuracy relations for real Roediger, H., & Geraci, L. (2007). Aging and the misinformation effect: A and suggested events. Memory, 8(5), 273–283. neuropsychological analysis. Journal of Experimental Psychology: Underwood, B. J. (1965). False recognition produced by implicit verbal Learning, Memory, and Cognition, 33(2), 321–334. doi:10.1037/ responses. Journal of Experimental Psychology, 70, 122–129. 0278-7393.33.2.321 Wade, K. A., Garry, M., Don Read, J., & Lindsay, D. S. (2002). A picture is Roediger, H. L., & McDermott, K. B. (1995). Creating false memories: worth a thousand lies: Using false photographs to create false child- Remembering words not presented in lists. Journal of Experimental hood memories. Psychonomic Bulletin & Review, 9(3), 597–603. Psychology: Learning, Memory, and Cognition, 21(4), 803–814. doi:10.3758/BF03196318 Roediger, H. L., & McDermott, K. B. (2000). Tricks of memory. Current Watson, J. M., Bunting, M. F., Poole, B. J., & Conway, A. R. A. (2005). Directions in Psychological Science, 9(4), 123–127. doi:10.1111/ Individual differences in susceptibility to false memory in the 1467-8721.00075 Deese-Roediger-McDermott paradigm. Journal of Experimental Roediger, H., Watson, J., McDermott, K., & Gallo, D. (2001). Factors that Psychology: Learning, Memory, and Cognition, 31(1), 76–85. determine false recall: A multiple regression analysis. Psychonomic Wilkinson, C., & Hyman Jr., I. E. (1998). Individual differences related Bulletin & Review, 8(3), 385–407. to two types of memory errors: Word lists may not generalize Sanford, L. C., & Fisk, J. E. (2009). How does the extraversion personality to autobiographical memory. Applied Cognitive Psychology, 12 trait influence false recall with the Deese–Roediger–McDermott (7), S29–S46. doi:10.1002/(SICI)1099-0720(199812)12:7<S29::AID- (DRM) paradigm? Journal of Research in Personality, 43(6), 972– ACP596>3.0.CO;2-R 977. doi:10.1016/j.jrp.2009.05.011 Winograd, E., Peluso, J. P., & Glover, T. A. (1998). Individual differences Schacter, D. L., Buckner, R. L., Koutstaal, W., Dale, A. M., & Rosen, B. R. in susceptibility to memory illusions. Applied Cognitive Psychology, (1997). Late onset of anterior prefrontal activity during true and 12(7), S5–S27. false recognition: An event-related fMRI study. NeuroImage, 6(4), Wootan, S. S., & Leding, J. K. (2015). Need for cognition and false 259–269. doi:10.1006/nimg.1997.0305 memory: Can one’s natural processing style be manipulated by Spanos, N. P. (1996). Hypnosis, age regression, and memory. In N. P. external factors? The American Journal of Psychology, 128(4), 459– Spanos (Ed.), Multiple identities & false memories: A sociocognitive per- 468. doi:10.5406/amerjpsyc.128.4.0459 spective (pp. 91–104). Washington, DC: American Psychological Wright, D. B., & Loftus, E. F. (1999). Measuring dissociation: Comparison Association. Retrieved from https://vpn.nacs.uci.edu/+CSCO of alternative forms of the dissociative experiences scale. Australian +ch756767633A2F2F63666C706172672E6E636E2E626574++/books/ Journal of Clinical & Experimental Hypnosis, 28(2), 103–126. 10216/007 Wright, D. B., Startup, H. M., & Mathews, S. A. (2005). Mood, dissociation Spanos, N. P., Burgess, C. A., Burgess, M. F., Samuels, C., & Blois, W. O. and false memories using the Deese-Roediger-McDermott pro- (1999). Creating false memories of infancy with hypnotic and non- cedure. British Journal of Psychology, 96(3), 283–293. doi:10.1348/ hypnotic procedures. Applied Cognitive Psychology, 13,201–218. 000712605X49916 984 R. M. NICHOLS AND E. F. LOFTUS

Zaragoza, M. S., & Lane, S. M. (1998). Processing resources and eyewit- Zhu, B., Chen, C., Loftus, E. F., Lin, C., He, Q., Chen, C., … Dong, Q. ness suggestibility. Legal and Criminological Psychology, 3(2), 305– (2010a). Individual differences in false memory from misinforma- 320. doi:10.1111/j.2044-8333.1998.tb00368.x tion: Personality characteristics and their interactions with cognitive Zaragoza, M. S., Lane, S M., Ackil, J. K., & Chambers, K. L. (1997). abilities. Personality and Individual Differences, 48(8), 889–894. Confusing real and suggested memories: Source monitoring and doi:10.1016/j.paid.2010.02.016 eyewitness suggestibility. In N. L. Stein, P. A. Ornstein, B. Tversky, Zhu, B., Chen, C., Loftus, E. F., Lin, C., He, Q., Chen, C., … Dong, Q. (2010b). & C. Brainerd (Eds.), Memory for everyday and emotional events Individual differences in false memory from misinformation: Cognitive (pp. 401–425). Mahwah, NJ: Erlbaum. factors. Memory, 18(5), 543–555. doi:10.1080/09658211.2010.487051 Zhu, B., Chen, C., Loftus, E. F., Lin, C., & Dong, Q. (2013). The relationship Zoellner, L. A., Foa, E. B., Brigidi, B. D., & Przeworski, A. (2000). Are trauma between DRM and misinformation false memories. Memory & victims susceptible to “false memories”? Journal of Abnormal Cognition, 41(6), 832–838. doi:10.3758/s13421-013-0300-2 Psychology, 109(3), 517–524. doi:10.1037/0021-843X.109.3.517